Adlard E.R. (ed.) Chromatography in the Petroleum Industry

Подождите немного. Документ загружается.

Chapter

ble with the column packing and the chromatography conditions. Sample mole-

cules that are larger than the pore size

the stationary phase pass through the

column at the same rate as the solvent. Smaller molecules diffuse into the pores

and are retained longer on the column; the smallest molecules being retained for

the longest time. Usually several columns are used to obtain the required resolu-

tion. They are packed with either mixed pore size material,

each column con-

tains a different pore size packing to cover the molecular range

the sample

components. The calibration and data handling have to be specially designed to

produce a molecular weight distribution curve that is the conventional way

round.

There are many books on

SEC

in general, but only

few which relate to low

molecular weight polymers such as waxes

[86,87].

These books adequately

cover the theory, equipment and general practice

SEC.

Therefore, only the

relevant points relating

wax chromatography are mentioned.

SEC

is a low

resolution separation method.

For

most hydrocarbon waxes,

produces a chro-

matogram consisting of a single curve with no discrete peaks (except for mixed

alkane standards where single peaks will be seen). The curve envelope contains

the sum of all the branched and straight chain alkanes merged together in

dis-

tribution from low to high molecular weight. The resolution

the curve depends

mainly on the size exclusion separation process, but it is also influenced by dis-

persion processes within the chromatography equipment (molecular diffusion,

column packing, etc.).

3.4.1

Early

work

the SEC analysis

waxes

The standard procedures

for

analysing waxes using

SEC

were established be-

fore

1980.

few years after the establishment of the technique in

1964,

it was

used to investigate cross-linked products from the irradiation

long-chain al-

kanes

[88,89].

Salovey

al.

carried out an excellent piece

research using a

series of polystyrene gel columns at

135"C,

with trichlorobenzene mobile phase,

to separate alkanes up to

C72.

It is unfortunate that their alkane calibration up to

C94

(including response factors) cannot be repeated currently, due to lack of

synthesized standards.

I97

Hillman produced an important comparative paper on the

GPC

and

GLC

separation

waxes

[90].

He used a variety of Styragel columns and sol-

vents, with refractive index detection. Tetrahydrofuran

(THF)

was used at

30°C

separate paraffin waxes, although these materials are more suitably analysed

GLC.

The

detector response

for

alkanes, with this solvent, varies widely

and at

C10

the response inverts

[91].

THF has a maximum operational tempera-

ture of

4OoC,

at which it requires careful handling (due to its toxic and unstable

nature), and

inappropriate

for

separating microcrystalline and synthetic waxes.

The chromatographic analysis

refined and synthetic waxes

Toluene was used at

70°C

for separating all waxes, but was found to be insensi-

tive when combined with the

detector for polyethylene waxes. However, at

that time, the preferred mobile phase was ortho-dichlorobenzene, combined with

an inhibitor, which was used at

80°C.

At high temperature, the chlorobenzenes

are highly toxic, therefore the equipment has to be placed in a fume cupboard

and the solvent has to be disposed of as’ toxic waste. The carbon number range

and maximum carbon number for both the

GPC

and

GLC

analysis of paraffin

and microcrystalline waxes were very similar. Several other papers on the

GPC

separation of waxes were written around this time, but they were all based on the

previous papers.

Before the 1980s, porous silica was used, but became unpopular due to the

possibility of non-size exclusion interactions taking place. The Styragel columns

used by HiIIman and others were long and wide

(120

7.8

mm). They

contained a large particle sized packing

(37-75

pm diameter) which gave typical

efficiencies of

3000

plates. This was equalled later by shorter columns

(30

long) containing smaller, 10 pm u-Styragel. These columns were coupled to-

gether in a carefully selected combination (or “column set”) of

60,

100, 200,

400,500

and

1000

pore size columns. The selection of each pore size depends

on the wax being analysed and the linearity required for the calibration curve

obtained by combining the columns. For the analysis of waxes, the columns and

injector should be placed in a thermostated oven. A long length of stainless steel

tubing has to be placed in the oven, between the pump and the injector, to pre-

heat the mobile phase.

The component parts of the

SEC

system were usually connected with the

minimum amount of stainless steel tubing (usually 0.25 mm bore). As mentioned

above, ortho-dichlorobenzene (with an antioxidant) was the preferred solvent for

wax analysis, because of its solvation power and compatibility with columns at

high temperature. The solvent was normally pumped at a

flow

rate of

ml min-I,

using either a syringe or a reciprocating pump to give a constant, pulse-free flow.

Before 1980 it was quite common to inject large volumes (up to

ml) onto the

column set via a septum, even although this increases the mobile phase viscosity

and sample band width. With the reduction in column particle size, injections

between

and 2000pl were made using four port, loop injectors. Automated

injection valves had been introduced by this time but were not commonly used.

Automation and a small sample size reduces band spreading.

The most common detector used in this work was the refractive index

(RI)

detector. As electronics and other ideas were developed over the years,

the

detector was improved. However, it lacks sensitivity for waxes and its response

for a homologous series of alkanes can vary dramatically for some solvents

(even going negative) [91]. Also as the working temperature of the

detector is

raised, its instability increases. In 1975, Dawkins and Hemmings used a modi-

References pp.

90-93

Chapter

fied infrared detector, with

heated cell, for the high temperature analysis of

polyolefins [92]. Infrared detection is limited by lack of sensitivity and by the

shortage of suitable solvents with an absorbance window at the appropriate

wavelength to detect hydrocarbons. During this period, several experimental de-

tectors were used;

these included the Pye Unicam moving belt and

FID

“transport detector”, and the evaporative light scattering detector

(ELSD),

evaporative mass detector which was introduced in 1966 [93].

The detector signal was generally sent to a chart recorder and an integrator.

The integrator would be programmed to carry out a peak area slice routine and

calculate the mean molecular weight. Most

SEC

researchers in the oil industry

only had access to these basic tools, and calibration had to be carried out

manually. However,

few

SEC

research teams in other areas did have access to

mainframe computers to calculate the baseline drift, carry out area counts, store

calibrations and calculate the separate molecular weight parameters. It is fortu-

nate for wax chromatographers that calibration can be carried out using stan-

dards of similar molecular size. The pure monodispersed alkanes, and Polywaxes

very narrow molecular mass distribution were well used as calibrants during

this period. The use of highly characterized broad molecular mass calibrants was

introduced by the

National Bureau of Standards (NBS) in the early 1970s.

However,

was not until 1978 that a low molecular weight, linear polyethylene

standard

(NBS

SRM

1482) was produced covering the upper range of synthetic

waxes [94].

3.4.2

Present

day

technology

The precise electronic control of pumps, ovens and detectors, plus advances in

detector design, have made

SEC

data more accurate and reproducible. Develop-

ments

column technology (i.e. the reduction in particle size and column di-

ameter, plus the development of mixed pore size packings) have enabled waxes

to be separated efficiently. Data handling, using powerful

PCs

and sophisticated

software, has allowed the analyst to measure and view the data in far greater de-

tail than previously.

Also,

after a moribund period between the mid-1970s and

mid-l980s, the status quo of high temperature

SEC

and low molecular weight

polymer analysis

now being questioned. There is a definite need for high

resolution SEC analysis of waxes, particularly for characterizing new synthetic

waxes and for investigating mixed polymer blends containing waxes (e.g. hot

melt adhesives).

1980,

the use of

series

25-30 cm long columns,

mm in internal di-

ameter, was common. Each column contained 10pm diameter particles of dif-

ferent pore size. The columns were arranged to give the best calibration for

log

molecular weight versus retention time graph (typically

third order polynomial

The chromatographic analysis

refined and synthetic waxes

calibration). In 1982, Waters introduced the Ultrastyragel columns with three

times the efficiency of the u-Styragel columns By the mid-l980s,

pm mixed

bed columns (each containing all the pore sizes used in a column series) were

commonly used in tandem to separate polyolefins. However, these mixed bed

columns did not have the resolution appropriate for microcrystalline and syn-

thetic wax analysis. Also the high temperature polyolefin analytical method

(using di- or tri-chlorobenzene eluent) is not appropriate for waxes or wax

blends, as the antioxidant peak in the eluent interferes with the low molecular

weight end of the chromatogram [95].

Following these improvements in packing materials, reliable

pm polysty-

rene-divinylbenzene phases, for individual pore size columns or mixed gel col-

umns, became available for low molecular weight analysis [96]. These columns

are much more efficient and have a higher resolution than the 10pm columns.

The Polymer Laboratories 5pm Mixed-D column has an efficiency

>50

000

platedm and is linear over the molecular weight range 200-400

000.

However, it

is important to reduce dead spaces to a minimum, e.g. by reducing connecting

tubing internal diameter and length, or the improvements will be lost in band

broadening. It is also important to use a guard column of the same material to

avoid column blockage at the injection end. Mixed bed columns can be com-

bined together to obtain higher resolution, without the calibration discontinuities

that occur when combining several different pore size columns.

In 199 1 Polymer Laboratories introduced a

pm mixed pore size column [97]

for separating low molecular weight materials (with a linear calibration up to

000

molecular weight). These columns have an exceptional efficiency

(approaching 100

000

plates/m) and can be used for rapid analysis. However,

they can only be used up to 100°C and require a system with low dead volume

and an accurate micropump to work effectively. Low dead volume is achieved

by using short, narrow connecting tubing and a low volume injector (i.e. lop1 or

less).

useful advantage is the lower solvent consumption, but the low flow rate

can lead to detection problems.

work has yet been published on the use of

these columns for wax analysis.

Several authors have tried to use packed SEC material in long capillary col-

umns [98]. MicroSEC, as it is called, is a very difficult technique and typically

can give an efficiency up to

000

platedm [99] (similar to the standard

columns). These columns are also much more expensive than standard columns

[99]. The more practical options are the 4.6 mm i.d. columns (produced by Phe-

nomenex and Polymer Laboratories) and the 2.1 mm i.d. columns (produced by

Phenomenex). These columns are packed with

pm mixed pore size material.

Ying

al.

[

1001 used cyclohexane at 70°C as a mobile phase to separate iso-

tactic polypropylene (after it had been dissolved in hot decahydronaphthalene to

destroy the crystallinity). This is a much less toxic material than chlorobenzenes

References pp.

9&93

Chapter

lfi

and has a lower viscosity. It has been claimed

[87]

that aliphatic hydrocarbons

shrink the column packing, causing voids and channels to form, and lowering the

efficiency. Ying

al.

found to the contrary. Their u-Styragel columns were used

continuously with cyclohexane for more than

months, with no change in reso-

lution. Since then, the stability of styrene-divinylbenzene columns

high tem-

perature

has

been much improved.

Barker has used cyclohexane at 70°C in

series of

pm columns to success-

fully separate both refined and synthetic waxes

[95]

(see Fig.

3.9).

This solvent

is difficult to

use

in conjunction with a refractive index detector, because the

refractive index difference between cyclohexane and waxes is very small and

therefore the sensitivity is low. Nor can it be used with an infrared detector as

there is no appropriate solvent window. However, it can easily be used, in con-

junction with the evaporative mass detector, to detect

all

waxes with higher

sensitivity.

The evaporative mass detector (or

ELSD)

is manufactured by several compa-

nies worldwide (Polymer Laboratories Ltd,

UK;

Varex Corporation,

USA;

Sedex, France). Each

different design, but they all work on the same princi-

ples

[loll.

The column effluent is passed through

nebulizer and converted into

0.0

0.04

0.05

0.06

0.07

0.08

0.09

0.10 0.11

0.12

Molecular

Weight

10e4

Fig.

3.9.

Overlaid

SEC

analysis

refined and synthetic waxes.

Polywax

500;

----,

mi-

crocrystalline

wax;

Polywax

655.

The chromatographic analysis

refined and synthetic waxes

a fine spray using pressurized nitrogen. The spray passes along a heated column

where the solvent is volatilized at a controlled temperature, leaving the non-

volatile solute to pass through a light beam (produced by a filament lamp or la-

ser). Light is scattered by reflection and refraction off the separated component

particles and is detected at

120”

to the light source by a photomultiplier or

photodiode detector. If the experimental condition

are

optimized

[

102,103], the

output from the detector is proportional to the mass of eluent over the whole of

the molecular weight range. This makes it a sensitive, universal mass detector.

However, it is essential that the connection between the column oven and the

detector is as narrow and as short as possible. This tube should also be heavily

insulated and preferably heated, to minimize the “plating out” of the high mo-

lecular weight components on the inner walls.

Until recently, ELSD instruments did not have the same response factor for

alkanes over the whole of molecular weight range, with a tailing off of the re-

sponse for lower alkanes (due to their volatility). Also the signal of each compo-

nent may not be directly proportional to the concentration. Varex have produced

electronic “linearizer” to overcome the latter effect for single component

peaks. Recently, Polymer Laboratories Ltd. have redesigned the evaporative

mass detector (Model PL-EMD

950)

to address the problems inherent in all

previous instruments of non-linear response and concentration effects and have

greatly increased the sensitivity of the detector. This has been mainly achieved

by placing two fibre optic bundle collectors at an optimum angle of

120”

either

side of the light source. They are connected to a remote photomultiplier tube.

Also

a heated connection with the column oven has been produced. The detector

works best if the sample concentration is small (i.e.

0.001-0.20%

of solute) and

the solvent evaporation temperature is as low as possible (vaporization depends

on latent heat not boiling point). These conditions are ideal for waxes, particular

the higher molecular weight synthetic materials which, at room temperature, tend

to be insoluble in many solvents.

A more sophisticated and expensive form of the ELSD is the multi-angle laser

light scattering detector (MALLSD). The column effluent is examined directly

in a specially designed flow cell. A laser light beam passes through the cell and

the scattered light is detected by an array of photodiode detectors arranged in a

circle either side of the light path. Using this arrangement, it is possible to meas-

ure absolute molecular weight and size distribution accurately. This detector was

originally restricted to higher molecular weight polymer analysis, but Wyatt

Technology Corporation, USA have now proved that it is possible to success-

fully use it for high temperature SEC and low molecular weight samples (i.e.

polystyrene molecular weight

760)

[

1041. However, it requires a concentration

detector in series as the sensitivity of MALLSD increases with higher molecular

weight. With waxes, this wouId normally be the refractive index detector, as a

References

pp.

90-93

Chapter

high sample concentration is required for the

MALLSD.

However, refractive

index is an inaccurate means of mass detection at low molecular weight due to

the changing response factor of the components.

Another detector which could be used is the viscometer, but this also requires

a concentration detector (usually refractive index). Viscosity detection has been

used since the early 1970s

[105].

It was not until the mid-1980s that

sensitive

differential viscosity detector, that could be used with SEC, was produced

Viscotek,

USA

[106].

In SEC, the molecules are separated by molecular size, i.e.

the hydrodynamic volume. This

equal to the sum of the intrinsic viscosity and

the molecular weight of the polymer. The calibration curve for the logarithm of

the hydrodynamic volume versus retention volume of all polymers

the same.

Therefore, with this detector it is possible to produce a universal calibration and

accurately measure the molecular weight

an unknown. This detector can also

be used in conjunction with known standards to plot

graph of branching fre-

quency versus molecular weight. Although the differential viscometer has been

used extensively

for

polyethylene analysis, it does not appear to have been used

for

waxes,

The

mathematical relationship of intrinsic viscosity and hydrody-

namic volume with low molecular weight materials is more complex than at

higher molecular weights

[

1071. Also refractive index is an inaccurate means of

mass detection at low molecular weight. Therefore care is needed in using this

detector for analysing waxes.

Other detectors have been developed, but do not appear to be commercially

popular. In

1983,

a flame ionization detector

for

liquid chromatography was pro-

duced by Tracor Instruments,

USA [108].

This should have a linear mass re-

sponse for all alkanes. A thermospray interface coupled to a Fourier transform

infrared detector by a moving metal ribbon [lo91 shows promise, but has not

been commercialized.

The most promising recent innovation has been an infrared detection system

produced by Lab Connections Inc.

lo]. This involves spraying the SEC eluent

onto the edge of

rotating germanium disc, where the solvent

removed, and

then analysing the deposited material by rotating the disc in an infrared spec-

trometer. Lab Connections have used this detector for high temperature SEC

analysis of low density polyethylene and

for

determining branching as a function

molecular weight distribution. It shows great promise for wax blend analysis

where additives such as ethylene vinyl acetate are part of the composition. An-

other promising technique that

being increasingly used for LC separation, but

does not appear to have been used in SEC, is electrospray mass spectrometry.

Combined with high resolution microbore

SEC,

it should be possible to obtain

absolute data on molecular weight distribution and branching.

There have been great advances in data collection from detectors, PC hard-

ware capability and data handling software. This has opened up access to expert

The chromatographic analysis

refined and synthetic

waxes

systems for SEC data handling

[

1 1

11. Polymer Laboratories have been the leader

in using Microsoft Windows software to enable the analyst to write his own

sample data handling method, calibrate his instrument and view the results more

easily than before

[

121. Fluctuations in flow rate between calibration and sam-

ple runs can be offset using internal standardization techniques, and the chroma-

togram is corrected accordingly. The analyst can now carry out post-analysis

examination, including overlaying chromatograms and exporting the data to a

spreadsheet

[

121.

Calibration procedures have been clearly dealt with in publications already

cited

[87].

1968,

was possible to carry out an SEC calibration using alkanes

up to molecular weight 1500 [88]. Today, only alkanes

to molecular weight

962 are available. Polymer Laboratories produce a series of Polywaxes and

polyethylene standards up to 14

000

molecular weight which can be used for

narrow range calibration of synthetic waxes. However, some are not ideal be-

cause they have a polydispersity

(M,

M,)

greater than 1.1. The author is at

present investigating the use

new synthetic waxes to extend the molecular

weight range for narrow standards. In the United States, NBS Standard

SRM

1482 has been well characterized

a low molecular weight broad standard [94].

Broad standards are not readily available, as they require fractionation and char-

acterization of the components to produce a cumulative weight fraction table.

They are a single run standardization and are valid for non-linear calibrations.

With the advent

high quality mixed bed columns and detectors, plus the use

accurate routine calibration, internal standardization and modem data han-

dling, it

now possible to obtain accurate and reproducible SEC chroma-

tograms. Apart from the standard log. molecular weight distributions, it is now

possible to view the linear molecular weight distributions (dwldMversus

and

the cumulative linear molecular weight distributions

[95].

This is very relevant

to the

SEC

waxes where the molecular weight distribution is relatively narrow

(see Fig. 3.9). It now appears possible to carry out a quantitative

SEC

mass dis-

tribution analysis for waxes, similar to the GC carbon number distribution

analysis (but not with the same precision).

3.5

CONCLUSIONS

Instrumental and computer advances over the past 10 years have given the

analyst the tools to investigate waxes and wax blends in great detail. It is now

possible to use GC to clearly separate waxes into straight and branched chain

alkanes

for

quantitative analysis. However, there is still a need for a thorough

investigation into the correct method

peak integration. Hopefully the stan-

dardization bodies will resolve this issue in the near future. Harmonization be-

References

pp.

90-93

Chapter

tween wax blend producers and customers would then be easier and product de-

velopment more rapid.

SFC separation has not been popular, but

SFE

could prove useful for trace

additive and contamination analysis in waxes (particularly with increasing hy-

giene legislation). Recent advances in

SEC

column technology, detector design

and computer software have made it possible to obtain an accurate molecular

weight distribution

waxes, particularly when they are blended with other

polymers. However, there is a need for more standards between n-C60 and the

highIy characterized

NBS

polyethylene standards. Accurate

SEC

analysis of raw

materials and their blends, combined with other methods of analysis (e.g. differ-

ential scanning calorimetry and rheology), could be developed into a powerful

research tool. This would enable the materials scientist to understand raw mate-

rials and predict properties of blends more accurately.

3.6

REFERENCES

Jowett,

G.D.

Hobson (Ed), Modern Petroleum Technology, Pt. 2, Wiley, New

York

(1986)

1021.

T.R. Graves and

J.G.

McCormick, 1987 ASC Raw Materials Seminar, Cincinnati, OH, p. 69;

Adhesive Age, 31(7) (1988) 22.

Krutzel, Tappi 1988 Hot Melt Symposium Notes, p. 59.

O.H. Hsu and H.S. Bender, Ind. Eng. Chem. Res., 27 (1988) 1296.

M. Huntley, Adhesive Age, Nov. (1991) 28.

C.G.

Scott and

D.A.

Rowell, Nature 187 (1960) 143.

F.J. Ludwig, Anal. Chem., 37 (1965) 1732.

R.W. Finch, Analabs Research Notes, 10 (1970) (3)l.

A.P. Tulloch, J.

Am.

Oil Chem. SOC., 50 (1973) 367.

T.H.

Gouw,

I.M. Whittemore and R.E. Jentoft, Anal. Chem. 42 (1970) 1394.

Proposed

Test

Method

for

Carbon Number Distribution

Paraffin Wax by

Gas

Chromatog-

raphy, ASTM, November (1 979).

lnstitute

Petroleum Method, Designation 372/85

985).

Barker, in G.B. Crump (Ed), Petroanalysis’87, Wiley,

New

York (1988), pp. 159-171.

R. Dandeneau and E.H. Zerenner, J. HRC and CC, 2 (1979) 351.

A.W. Aue and

P.P.

Wickramanayake, J. Chromatogr., 200 (1979) 3.

1,.

Blomberg, J. Buijten,

Markides and

Waenmann, J. Chromatogr., 208 (1981) 231.

LA.

Luke and

J.E.

Ray, J. High Resolut. Chromatogr. Column Chromatogr.,

(1985) 193.

Trestianu,

Zilioli, A. Sironi, C. Saravalle,

Munari, M. Galli, G. Gaspar, J. Colin and

Jovelin, J. High Resolut. Chromatogr. Column Chromatogr.,

(1985) 771.

Dawes and M. Cumbers, Int. Lab., 19 (1989) 9,34.

P. Sandra, Sample Introduction in Capillary

Gas

Chromatography, Vol. 1, Huethig Verlag

(1985).

Schomburg, H. Hussmann and R. Ritterman,

Chromatogr., 204 (1981) 85.

Schomburg and H. Hussmann, German Patent Application (1976).

W. Vogt,

Jacob and H.W.J. Obwexer, J. Chromatogr., 174 (1979) 301.

The chromatographic analysis

refined and synthetic waxes

K. Grob, Jr., G. Karrer and M.I. Riekkola, J. Chromatogr.,

334 (1985) 129.

J.M. Colin and J.L. Jovelin, Total-France, Personal Communication, July

(1987).

Poy, Demonstration at 4th Int. Symp. on Capillary Chromatography, Hindelang

(1981).

G. Schomburg, in R.E. Kaiser (Ed), Capillary Chromatography, 4th Int. Symp. Hindelang

1981,

Huthig Verlag

(1981),

921A.

W. Jennings and M.F. Mehran

Chromatogr. Sci.,

24 (1986) 34.

K. Grob,

Th.

Laubli and B. Brechbuhler, J. High Resolut. Chromatogr. Column Chromatogr.,

(1988) 462.

J.V. Hinshaw, J. Chromatogr. Sci.,

25 (1987) 49.

J.V. Hinshaw and L.S. Ettre, J. High Resolut. Chromatogr. Column Chromatogr.,

12 (1989)

251.

A. Barker, MSc. Thesis, University of Greenwich, London

(1987).

A. Tipler and G. Johnson, J. High Resolut. Chromatogr. Column Chromatogr.,

(1990) 365.

J. Hinshaw, LC-GC Int.,

5(12) (1992) 14.

H.-G. Janssen, Chromatography Technical Notes, Ai Cambridge Ltd.

David and P. Sandra, LC-GC Int.,

5(12) (1992) 22.

J.R. Berg and T. Hawkins, Int. Lab.,

9(5) (1992) 10.

J.V. Hinshaw, LC-GC Int.,

3(4) (1990) 34.

Coffey (Ed), Rodd’s Chemistry of Carbon Compounds,

2nd

edition, Vol.

lA,

Elsevier,

Amsterdam

(1964).

M.F.

Ansell (Ed), Supplement to Rodd’s Chemistry of Carbon Compounds, 2nd edition, Vol.

Elsevier, Amsterdam

(1975).

J. Szajranek, J. Kusmierz and W. Czerweic, J. Chromatogr.,

245 (1982) 219.

L.A. Luke and J.E. Ray, Proc. Chromatography Society, Bristol University

(1984).

Munari, C. Saravelle, A. Della Foglia,

Trestianu, M. Gallin, J. Colin and J. Jovelin, Proc.

6th Int. Symp. on Capillary Chromatography

(1985),

771.

R.D. Dandeneau and E.H. Zerenner, J. High Resolut. Chromatogr. Column Chromatogr.,

(1979)

351.

L. Blomberg and T. Wannman, J. Chromatogr.,

168 (1979) 81.

J. Barberio, Perkin Elmer

The Analytical Report No.

10 (1986)

S.R. Lipsky and M.L. Durn, J. High Resolut. Chromatogr. Column Chromatogr.,

9 (1986)

376.

S.R. Lipsky and M.L. Durn, LC-GC,

4 (1986) 898.

J. Zeevin, Chrompack U.K. Seminar- G.C. in the Petroleum Industry

(1987).

Y. Takayama, T. Takeichi and

Kawai, J. High Resolut. Chromatogr. Column Chromatogr.,

988) 732.

J. Buyten, C. Duvekot, J. Peene, J. Lips and

Mohnke,

lth Int. Symp. on Capillary Chro-

matography, Monterey

990),

Chrompack, Capillary Column for High Temperature Gas Chromatography, Reprint No.

503551 (1990).

Schuyler, J.W. Stauffer, C.E. Loope and C.R. Vargo, Process Control Quality,

(1992)

167.

D.H. Desty,

Goldup and W.T. Swanton, in N. Brenner, J.E. Callen and M.D. Weiss (Eds),

Proc. 3rd ISA Symp. on Gas Chromatography, Michigan State University,

1961,

Academic

Press, New York

(1962),

pp.

105-185.

C.P.M. Schutjes, E.A. Vermeer, J.A. Rijks and C.A. Cramers, J. Chromatogr.,

253 (1982)

Letter and data sheet from Dr. N. Douklais; MicroQuartz, Kistlerhofstrasse

243, D-8000

Munchen

70,

Germany

(1 989).