U.S. patent application number 14/065988 was filed with the patent office on 2014-07-10 for electro-spun fibers and applications therefore.
This patent application is currently assigned to THE OHIO STATE UNIVERSITY RESEARCH FOUNDATION. The applicant listed for this patent is THE OHIO STATE UNIVERSITY RESEARCH FOUNDATION. Invention is credited to Jonathan E. Clark, Susan V. Olesik, Jeremy K. Steach, Joseph W. Zewe.
Application Number | 20140193589 14/065988 |
Document ID | / |
Family ID | 41478987 |
Filed Date | 2014-07-10 |
United States Patent
Application |
20140193589 |
Kind Code |
A1 |
Olesik; Susan V. ; et
al. |
July 10, 2014 |
ELECTRO-SPUN FIBERS AND APPLICATIONS THEREFORE
Abstract
A supported nanofiber medium useful for segregating chemical
species is provided by selecting a polymer, selecting a substrate;
and electrospinning the polymer to form a nanofiber medium on the
supporting substrate. When the substrate is a planar surface, the
nanofiber medium will be a mat suitable for conducting
chromatographic separation. When the substrate is a filament, the
nanofiber medium is an annular mat suitable for solid phase
microextraction. The nanofiber media formed may be selectively
cross-linked and at least partially carbonized to carbon
nanofibers. The nanofiber medium is supported on the substrate
without the use of binder material.
Inventors: |
Olesik; Susan V.; (Dublin,
OH) ; Clark; Jonathan E.; (Nashville, TN) ;
Steach; Jeremy K.; (Kingsport, TN) ; Zewe; Joseph
W.; (Marion Center, PA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
THE OHIO STATE UNIVERSITY RESEARCH FOUNDATION |
Columbus |
OH |
US |
|
|
Assignee: |
THE OHIO STATE UNIVERSITY RESEARCH
FOUNDATION
Columbus
OH
|
Family ID: |
41478987 |
Appl. No.: |
14/065988 |
Filed: |
October 29, 2013 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13046552 |
Mar 11, 2011 |
8568587 |
|
|
14065988 |
|
|
|
|
PCT/US2009/056685 |
Sep 11, 2009 |
|
|
|
13046552 |
|
|
|
|
61096139 |
Sep 11, 2008 |
|
|
|
Current U.S.
Class: |
427/483 |
Current CPC
Class: |
B01D 15/20 20130101;
G01N 2030/062 20130101; G01N 2030/009 20130101; B01J 2220/54
20130101; D01D 5/0084 20130101; D01F 9/12 20130101; B01J 20/28033
20130101; G01N 30/93 20130101; B01J 20/3295 20130101; G01N 30/06
20130101; B01J 20/28007 20130101; B01J 20/286 20130101; B01J
20/28035 20130101; B01J 20/3293 20130101; B82Y 30/00 20130101; B05D
5/00 20130101 |
Class at
Publication: |
427/483 |
International
Class: |
B05D 5/00 20060101
B05D005/00 |
Goverment Interests
STATEMENT REGARDING FEDERALLY-SPONSORED R & D
[0002] The present invention was made with Government support under
at least one of the following grants awarded by the National
Science Foundation: 0616709, EEC-04-25626, and CHE-04-40499. The
United States Government may have certain rights to this invention
under 35 U.S.C. .sctn.200 et seq.
Claims
1. A method for manufacturing a supported nanofiber medium useful
for segregating chemical species, comprising the steps of:
selecting a polymer; selecting a substrate; and electrospinning the
polymer, forming a nanofiber medium on the supporting
substrate.
2. The method of claim 1, wherein: the substrate is a planar
surface; and the nanofiber medium comprises a mat suitable for
conducting chromatographic separation.
3. The method of claim 1, wherein: the substrate is a filament; and
the nanofiber medium is an annular mat suitable for solid phase
microextraction.
4. The method according to claim 1 wherein further comprising the
step of: selectively cross-linking the nanofiber medium.
5. The method according to claim 2 wherein further comprising the
step of: selectively cross-linking the nanofiber medium.
6. The method according to claim 3 wherein further comprising the
step of: selectively cross-linking the nanofiber medium.
7. The method according to claim 1 further comprising the step of:
selectively cross-linking the nanofiber medium.
8. The method according to claim 2 further comprising the step of:
selectively cross-linking the nanofiber medium.
9. The method according to claim 3 further comprising the step of:
selectively cross-linking the nanofiber medium.
10. The method according to claim 4 further comprising the step of:
selectively cross-linking the nanofiber medium.
11. The method according to claim 5 further comprising the step of:
selectively cross-linking the nanofiber medium.
12. The method according to claim 6 further comprising the step of:
selectively cross-linking the nanofiber medium.
13. The method according to claim 1 further comprising the step of:
heating the nanofiber medium sufficiently to convert at least a
portion of the nanofibers to glassy carbon nanofibers.
14. The method according to claim 2 further comprising the step of:
heating the nanofiber medium sufficiently to convert at least a
portion of the nanofibers to glassy carbon nanofibers.
15. The method according to claim 3 further comprising the step of:
heating the nanofiber medium sufficiently to convert at least a
portion of the nanofibers to glassy carbon nanofibers.
16. The method according to claim 4 further comprising the step of:
heating the nanofiber medium sufficiently to convert at least a
portion of the nanofibers to glassy carbon nanofibers.
17. The method according to claim 5 further comprising the step of:
heating the nanofiber medium sufficiently to convert at least a
portion of the nanofibers to glassy carbon nanofibers.
18. The method according to claim 6 further comprising the step of:
heating the nanofiber medium sufficiently to convert at least a
portion of the nanofibers to glassy carbon nanofibers.
19. The method according to claim 1, wherein the supported
nanofiber medium is devoid of binder material.
20. The method according to claim 2, wherein the supported
nanofiber medium is devoid of binder material.
21. The method according to claim 3, wherein the supported
nanofiber medium is devoid of binder material.
22. The method according to claim 4, wherein the supported
nanofiber medium is devoid of binder material.
23. The method according to claim 5, wherein the supported
nanofiber medium is devoid of binder material.
24. The method according to claim 6, wherein the supported
nanofiber medium is devoid of binder material.
25. The method according to claim 7, wherein the supported
nanofiber medium is devoid of binder material.
26. The method according to claim 8, wherein the supported
nanofiber medium is devoid of binder material.
27. The method according to claim 9, wherein the supported
nanofiber medium is devoid of binder material.
28. The method according to claim 10, wherein the supported
nanofiber medium is devoid of binder material.
29. The method according to claim 11, wherein the supported
nanofiber medium is devoid of binder material.
30. The method according to claim 12, wherein the supported
nanofiber medium is devoid of binder material.
31. The method according to claim 13, wherein the supported
nanofiber medium is devoid of binder material.
32. The method according to claim 14, wherein the supported
nanofiber medium is devoid of binder material.
33. The method according to claim 15, wherein the supported
nanofiber medium is devoid of binder material.
34. The method according to claim 16, wherein the supported
nanofiber medium is devoid of binder material.
35. The method according to claim 17, wherein the supported
nanofiber medium is devoid of binder material.
36. The method according to claim 18, wherein the supported
nanofiber medium is devoid of binder material.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a divisional of and claims priority to
U.S. patent application Ser. No. 13/046,552, filed Mar. 11, 2011,
now U.S. Pat. No. 8,568,587, issued Oct. 29, 2013, which is a
continuation-in-part of PCT/US2009/056685, filed Sep. 11, 2009,
which claims priority to U.S. Provisional Patent Application No.
61/096,139, filed Sep. 11, 2008, all of which are incorporated by
reference as if fully recited herein.
TECHNICAL FIELD
[0003] The disclosed embodiments are in the field of producing
fibers by an electrospinning process and in applying the fibers so
produced to industrially useful applications. In a first
application, the fiber is integrated into a stationary phase of a
thin-layer chromatography system. In a second application, the
fiber is integrated into a system for solid phase
microextraction.
BACKGROUND
[0004] Thin layer chromatography (TLC) is a technique known in
chemistry to separate chemical species, and has a wide range of
uses in the fields of pharmaceuticals, water quality analysis and
organic synthesis. A stationary phase provides a tortuous flow
path, and the thin layer is often formed from a material such as
silica gel, alumina or cellulose, with the material selected for
the porosity provided and its inert chemical nature, relative to
the components of the liquid phase. A liquid phase consists of a
mixture of compounds to be separated, dissolved in an appropriate
solvent. The liquid phase is placed on the thin layer plate and is
drawn up the plate by capillary action, with the mobility of the
compounds in the mixture determined by the interactions of the
analytes with the stationary phase.
[0005] Known impediments to wider use of the TLC technique include
the cost and difficulty of fabricating the stationary phase, as
well as the brittleness of it. A further impediment is the minimum
size of the diameter and thickness available for a stationary phase
plate fabricated from the conventional materials.
[0006] Solid phase microextraction (SPME) is an extraction
technique that was developed as a fast method to analyze volatile
and semivolatile compounds. SPME offers an inexpensive and
solvent-free extraction method that is a viable alternative to
liquid-liquid extractions and solid phase extractions. This
technique was developed by Berladi and Pawliszyn in 1989.
[0007] A typical apparatus used in SPME comprises a silica rod or
platinum wire that is coated with a specific extraction medium.
SPME fibers often have a length of about 1 cm, and the desired
thickness of the extraction medium ranges from a monolayer to about
100 .mu.m. The coated fibers are used by placing the fiber in
contact with the sample matrix, a bulk fluid or its surrounding
headspace. The fibers can be used to analyze samples from the
liquid phase, gaseous, or solid matrices. The samples are measured
by direct exposure of the fiber to the bulk fluid or its
surrounding headspace.
[0008] By developing a fiber coating with different
functionalities, SPME fibers can be chosen for the specific
analyte(s) of interest. This allows the performance of SPME to be
optimized where only the analytes of interest are analyzed rather
than having matrix effects interfering in the extraction. By
decreasing matrix effects, the detection limit of the analytes can
be lowered.
[0009] As described, the separation techniques of TLC and SPME have
known impediments.
SUMMARY OF THE INVENTIVE CONCEPT
[0010] These and other impediments of the prior art separation
techniques are solved at least in part by a method for
manufacturing a supported nanofiber medium useful for segregating
chemical species, comprising the steps of: selecting a polymer;
selecting a substrate; and electrospinning the polymer, forming a
nanofiber medium on the supporting substrate.
[0011] Some aspects of the method provide a device for analytically
separating chemical species, comprising a substrate and a polymeric
nanofiber medium formed on at least one surface of the substrate,
where the substrate is a planar surface; and the nanofiber medium
comprises a mat suitable for conducting chromatographic
separation.
[0012] Other aspects provide a device where the substrate is a
filament; and the nanofiber medium comprises an annular mat
suitable for solid phase microextraction.
[0013] Yet further aspects of the method, by providing the
supported nanofiber media, provide an improved method for
segregating at least two chemical species.
BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In addition to the features mentioned above, other aspects
of the inventive concept will be readily apparent from the
following descriptions of the drawings and exemplary embodiments,
wherein like reference numerals across the several views refer to
identical or equivalent features, and wherein:
[0015] FIG. 1 is a schematic of an electrospinning apparatus
used;
[0016] FIG. 2 is a set of scanning electron microscope (SEM) images
showing products produced when varying the concentration of SU-8 in
the solution;
[0017] FIG. 3 is a set of SEM images showing pyrolyzed and
unpyrolyzed fibers obtained;
[0018] FIG. 4 is a set of SEM images showing pyrolyzed and
unpyrolyzed beads obtained;
[0019] FIG. 5 is an SEM image of typical UV patterned electrospun
fibers;
[0020] FIG. 6 is a set of SEM images of the electrospun SPME
fibers, unprocessed and after processing at various temperatures,
A: unpyrolized, B: 400.degree. C., C: 600.degree. C., D:
800.degree. C.;
[0021] FIG. 7 is an extraction time profile for benzenoid compounds
using an SU-8 electrospun fiber-coated SPME wire;
[0022] FIG. 8 is an extraction time profile for benzenoid compounds
using an electrospun fiber-coated SPME wire processed at
400.degree. C.;
[0023] FIG. 9 is an extraction time profile for benzenoid compounds
using an electrospun fiber-coated SPME wire processed at
600.degree. C.;
[0024] FIG. 10 is an extraction time profile for benzenoid
compounds using an electrospun fiber-coated SPME wire processed at
800.degree. C.;
[0025] FIG. 11 is an extraction efficiency plot for benzene using
various SPME systems;
[0026] FIG. 12 is an extraction time profile for phenolic compounds
using an SU8 electrospun fiber-coated SPME wire;
[0027] FIG. 13 is an extraction time profile for phenolic compounds
using an electrospun fiber-coated SPME wire processed at
400.degree. C.;
[0028] FIG. 14 is an extraction time profile for benzenoid
compounds using an electrospun fiber-coated SPME wire processed at
600.degree. C.;
[0029] FIG. 15 is an extraction time profile for benzenoid
compounds using an electrospun fiber-coated SPME wire processed at
800.degree. C.;
[0030] FIG. 16 is an extraction efficiency plot for 4-nitrophenol
using various SPME systems;
[0031] FIGS. 17 and 18 are a set of SEM images (FIG. 29 being at a
higher magnification) of a plan view of an embodiment of a TLC
plate manufactured under the electrospinning technique;
[0032] FIG. 19 is an SEM image of an edge-on electrospun TLC
device;
[0033] FIG. 20 shows characteristic Raman spectra of electrospun
SU-8 samples pyrolyzed at 600.degree. C. (A) and 1000.degree. C.
(B);
[0034] FIG. 21 shows the migration distance of the mobile phase
versus the square root of the migration time;
[0035] FIG. 22 shows a comparison of electrospun glassy carbon UTLC
devices with electrospun PAN UTLC devices for the analysis of laser
dyes;
[0036] FIG. 23 shows retardation factors for replicate studies
with, n=7, for each laser dye;
[0037] FIG. 24 presents the efficiency obtained using the glassy
carbon UTLC devices in comparison to work by Khan et al.;
[0038] FIG. 25 shows the Rf values for each of the amino acids.
[0039] FIG. 26 demonstrates the separations obtained with the
electrospun glassy carbon UTLC devices and comparison to previously
published literature values;
[0040] FIG. 27A is a digital photograph of an electrospun carbon
UTLC device;
[0041] FIG. 27B is illustrations of an exemplary electrospun carbon
UTLC device;
[0042] FIG. 27C is illustrations of an exemplary electrospun carbon
UTLC device;
[0043] FIG. 28 is graphical comparison of mobile phase
velocities;
[0044] FIG. 29 is a graphical comparison of various carbon
devices;
[0045] FIG. 30 is a chromotogram of the separation of a mixture of
S640, Rh 610 P and KR on a 600.degree. C. glassy carbon UTLC device
with a developement distance of 1.5 cm;
[0046] FIG. 31 is a graphical comparison of the plate number, N,
obtained for the analysis of laser dyes on electrospun PAN UTLC
plate electrospun glassy carbon UTLC devices;
[0047] FIG. 32 is a graphical comparison of the resolution of a
ternary mixture of laser dyes using various UTLC devices;
[0048] FIG. 33 is digital image and chromatogram of the separation
of a mixture of lysine, threonine and phenylalanine on a
600.degree. C. galssy carbon UTLC plate with a developement
distance of 2.0 cm; and
[0049] FIG. 34 is a graphical representation of the resolution of a
ternary mixture of 3 essential amino acids: lysine, threonine and
phenylalanine.
DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENT(S)
[0050] To the extent that they compliment the specification and are
enabling of the disclosed embodiments, the following publications
are incorporated by reference as if recited in their entirety
herein: Clark, Jonathan E., Technique for Ultrathin Layer
Chromatography Using an Electrospun, Nanofibrous Stationary Phase,
Anal. Chem. 2009, 81, 4121-29; Clark, Jonathan E., a manuscript
entitled: Electrospun Glassy Carbon UTLC Devices, submitted to The
Journal of Chromatography A; and The Doctoral Dissertation of
Jeremy Steach entitled: The Development of Novel Phases with
Photoresist for Capillary Electrophoresis, Capillary
Electrochromatography, and Solid-Phase Microextraction, available
at the Ohio State University.
Electrospinning Fibers
[0051] Referring first to the formation of the fibers that are
useful in the TLC and SPME applications, among other applications,
electrospinning involves placing a high electric field between a
polymer solution and a conductive collector. This collector may be
comprised of many different materials such as metals, conductive
polymers or the like, and may take the form of a plate, a film, a
filament, a rod etc. When an electric field strong enough to
overcome the surface tension of the droplet is provided, a Taylor
cone is formed. Following the creation of the Taylor cone, fibers
are ejected toward the conductive collector. With this technique,
many different polymers and polymer blends can be used to generate
and spin fibers with various chemical compositions and to fabricate
mats comprising the fibers without the aid of binders.
[0052] There are advantages of electrospinning a photoresist
material, such as the high contrast epoxy based photoresist
available for Microchem, Corp., of Newton, Mass., under the
designation SU-8 2100, referred to hereinafter as "SU-8." By using
SU-8 photoresist to electrospin micro/nanofibers, the nonwoven
fiber mats can easily be patterned after the fibers are electrospun
allowing greater flexibility in patterning. Patterning of a
photoresist polymer such as SU-8 is generally accomplished by
selectively exposing sections of the fibers to UV light and then
removing uncrosslinked polymer from the substrate.
[0053] An apparatus used for electrospinning is shown in FIG. 1.
SU-8 solutions including 100%, 75, 70, 50, 25 and 0% (v/v) were
prepared using commercially available cyclopentanone. A Spellman
CZE 1000R high voltage power supply was used to supply voltages
from 5 to 15 kV. A Harvard Model 33 dual syringe pump was used to
control the flow rate of the SU-8 solutions. The flow rates were
varied from 0 to 2 mL/min. Cleaved silicon pieces from a silicon
wafer were used as the collector for the fibers. Electrospinning
experiments were performed while varying the following parameters:
SU-8/cylcopentanone (v/v) concentrations, voltage, flow rate, and
distance to determine the optimum parameters for production of
fibers. After each experiment, the fibers were exposed to a UV
light to crosslink the SU-8 fibers.
[0054] Samples were prepared in sets of two by this electrospinning
method. After electrospinning, both samples were exposed to UV
radiation for approximately five minutes. One sample was then kept
in a vial and the other was placed in a quartz tube furnace for
pyrolysis. Once the sample was placed in the furnace, a forming gas
mixture (95% N2 and 5% H2) was flowed through the quartz tube to
remove any oxygen. The furnace was ramped at 1.degree. C./min to a
final temperature of 800.degree. C.
[0055] Forming gas flowed continuously for the duration of the
pyrolysis. The final temperature was held for a minimum of 5 hours
before the pyrolysis was stopped. After the furnace was turned off,
the sample was allowed to cool to room temperature (25.degree. C.)
under a constant flow of forming gas.
[0056] To pattern the fibers, the SU-8 was first electrospun onto
the silicon collector as described above. The pattern was created
by using a lithographic mask or by printing the desired shape onto
a transparency. The lithographic mask or transparency was placed
onto the fibers and then the fibers were exposed to the UV source.
After exposing the fiber samples to UV radiation, SU-8 developer
was used to remove the fibers that were not crosslinked.
[0057] All electron microscopy images of the electrospun fibers
were taken using a Hitachi S-4300 scanning electron microscope.
Each SU-8 sample was sputter coated with gold prior to obtaining
the images, while the pyrolyzed fiber samples were not coated to
obtain images.
[0058] After the initial experiments, several dilutions between 75%
and 25% were attempted to further optimize the conditions for fiber
production. It was found that 75% and 70% SU-8 solutions resulted
in fibers for the various voltages (5-15 kV), distances (5-29 cm),
and flow rates (0-0.2 mL/min) examined. The other concentrations
did not produce fibers or produced fibers along with polymer beads.
As shown in FIG. 2, the 50% and 25% SU-8 concentrations yielded
beads at the same parameters as 75% and 70% SU8 concentrations.
Both 100% and 0% SU-8 concentrations resulted in no fibers or
particles on the collector. The SEM images shown in FIG. 2 were
obtained for the various % SU-8 concentrations at the same
electrospinning parameters to determine the effective concentration
range to generate fibers.
[0059] The effect of distance on the fibers produced was also
measured for both the 70 and 75% solutions. Fibers were produced
effectively at distances from 10 cm to 29 cm. The 75% solution
yielded larger fibers than the 70% for each distance, ranging from
1.08 .mu.m at 10 cm to 830 nm at 29 cm. The 70% yielded average
fiber sizes ranging from 770 nm to 530 nm on average.
[0060] To test the effect of voltage on both solutions, the
distance was set to 10 cm and the flow rate was kept constant at
0.02 mL/min. For both 75% and 70% concentrations, the fibers
decreased in average diameter at certain voltages. For the 70%
solution, voltages below 9 kV were tested and did not provide any
smaller diameter fibers but actually the fiber size increased
slightly. At 5 kV, the 70% electrospun fibers were measured to be
830 nm.+-.340 nm compared to 490 nm.+-.140 nm for 9 kV. For the 75%
SU-8 dilution, the average fiber diameters were measured displayed
an even larger decrease in size when electrospun at 9 kV. The
average fiber diameter decreased from 1.08 .mu.m (10 kV) to 301 nm
(9 kV). Lower voltages were tested for the 75% concentration, and
the average fiber diameter increased compared to 9 kV. For example,
at a voltage of 5 kV, the average fiber diameter was 560 nm.+-.300
nm which is a slight increase in fiber diameter at the chosen
electrospinning parameters. So, a voltage of 9 kV was chosen to be
the best voltage for both concentrations under the electrospinning
parameters investigated.
[0061] Pyrolysis of electrospun SU-8 fibers generates carbon
fibers. FIG. 3 shows the pyrolyzed fibers produced from both the
75% and 70% SU8 solutions. Samples were pyrolyzed using
substantially the same method described above. The fiber diameter
decreased after pyrolysis but still maintained the same fiber
structure that existed prior to heating.
[0062] An advantage to electrospinning a polymer that can be
selectively cross-linked, such as SU-8 photoresist, is that
electrospun fiber patterns can be generated. FIG. 5 is an SEM image
of a pattern generated by exposing portions of the fiber mat to UV
light after electrospinning
[0063] In conventional electrospinning, the fibers are typically
patterned by using different collectors. Though these patterns
generate aligned or oriented fibers, most of them do not provide a
method to pattern the nonwoven fibers into several different
patterns. The advantage of being able to pattern fibers via UV
radiation is that UV photolithography can be used to easily
generate any fiber pattern. The ease of creating a patterned mat of
SU-8 fibers can lead to many application possibilities.
[0064] From this, it is shown that an optimized process exists to
generate electrospun fibers from an SU-8 2100 negative photoresist.
The electrospun SU-8 fibers generated had diameters ranging from
300 nm to 1 .mu.m which are dependent upon the set of parameters
used. The fibers may be easily patterned by UV and then converted
to carbon fibers via pyrolysis. The fiber structure is maintained
after pyrolysis. By utilizing the advantage of patterning the
fibers after electrospinning, many applications ranging from
microfluidics, sensors to micro/nanoelectronics can potentially be
achieved.
Solid Phase Microextraction
[0065] In one embodiment, electrospinning is employed to coat a
substrate, such as a stainless steel wire or other conductive
filament or surface, with a mat of fibers. This electrospinning of
polymeric fibers provides the capability to create micro/nanofibers
for chemical extraction and chromatography through an inexpensive
and simple method.
[0066] The parameters for the optimization of electrospinning SU-8
are described above. SPME fibers may also be pyrolyzed to yield
carbon-based fiber coated SPME wires. The extraction
characteristics of the SU-8 and pyrolyzed electrospun-coated wires
were investigated for nonpolar and polar compounds under headspace
extraction conditions.
[0067] Benzene (99.9%), toluene (99.8%), ethylbenzene (99.8%),
o-xylene (98%), phenol, 4-chlorophenol and 4-nitrophenol were used
for extractions. Dichloromethane was used to prepare the solutions
for benzene, toluene, ethylbenzene, and o-xylene, while HPLC
methanol was used to prepare the phenolic solutions.
[0068] For comparison, commercially available SPME fibers were
purchased from Supelco. The fibers chosen for comparative analysis
of nonpolar compounds were an 100-.mu.m polydimethylsiloxane (PDMS)
and a Stableflex 65-.mu.m polydimethylsiloxane/divinylbenzene
(PDMS/DVB). To compare the electrospun fiber coated wires for the
extraction of phenols; an 85-.mu.m polyacrylate (PA) commercial
fiber was used.
[0069] The 75% SU-8 2100/cyclopentanone electrospinning solution
was prepared as described above. Stainless steel wires were used as
the collector for the electrospun fibers. The wires had a diameter
of about 127 .mu.m and were commercially available from Small Parts
(Miami Lakes, Fla.).
[0070] The stainless steel wires were cut into 1.5 cm lengths and
previously determined parameters to electrospin a 75% SU-8 2100
solution were used to generate the fibers. The parameters for
electrospinning were: voltage of 9 kV, a flow rate of 0.02 mL/min,
and a distance of 10 cm. The fibers were electrospun for 30 seconds
to create the fiber coating on the stainless steel wire. After
electrospinning onto the stainless steel fiber, the fibers were
then exposed to a UV light to crosslink the SU-8 fibers.
[0071] Utilizing the previously described pyrolysis technique,
temperatures of 400, 600, and 800.degree. C. were chosen for
production of carbon fibers. The electrospun fiber-coated wires
were then transferred to the inlet on the gas chromatograph (GC).
The coated wires were held in the inlet for a minimum of 30 minutes
or until a steady baseline was achieved.
[0072] Extraction time profiles were initially determined by using
a mixture of benzene, toluene, ethylbenzene, and o-xylene (BTEX). A
40 ppm BTEX solution was prepared by spiking 25 mL of nanopure
water with a 10000 ppm BTEX solution in dichloromethane. After 20
minutes of stirring, the fiber was exposed to the headspace for a
set of selected extraction times (0 to 90 minutes) to determine the
extraction time profile.
[0073] A 10000 ppm BTEX solution was prepared in dichloromethane
and used to spike 25 mL of water to a concentration of 40 ppm. An
extraction time of 60 minutes was chosen to compare the extraction
efficiency of the electrospun fiber-coated wires to commercial
fibers.
[0074] The extraction time profiles for the phenolic compounds were
determined by using a mixture of phenol, 4-chlorophenol, and
4-nitrophenol. A 100 ppm solution of the above phenolic compounds
was prepared in HPLC methanol. To perform the extractions, the pH
of the nanopure water was lowered to .about.1 by the addition of
concentrated hydrochloric acid, while sodium chloride (NaCl) was
added to the water until saturated. This solution was spiked with
the 100 ppm phenolic mixture to give a concentration of 0.4
ppm.
[0075] To compare the extraction efficiency of each fiber, a 100
ppm phenolic solution was used. A time of 60 minutes was chosen as
the extraction time to compare the electrospun fiber-coated wires
to the PA commercial SPME fiber. For analysis of the extraction
efficiency, the inlet temperature was 320.degree. C. for the
electrospun-coated fibers and the inlet temperature for the PA
fiber was held at 280.degree. C. All other chromatographic
parameters were kept the same.
[0076] By coating the stainless steel wire with electrospun fibers,
a porous and high surface area coating is provided for solid-phase
microextraction. The nonwoven mat of electrospun fibers was formed
around the stainless steel wire collector. Electrospinning times
can be altered to yield different fiber coating thicknesses and or
density on the stainless steel wire for SPME.
[0077] FIG. 6 shows SEM images of electrospun SU-8 fibers on a
stainless steel wire, where the fibers were electrospun for 30
seconds. Also, FIG. 6 shows the electrospun fiber-coated wires
after being processed at the different temperatures (400, 600, and
800.degree. C.). The thickness of each fiber-coated wire was
measured by SEM and shown in the following table, where it is shown
that the fiber coating thickness decreased as the processing
temperature increased.
TABLE-US-00001 Fiber Coating thickness (.mu.m) SU-8 2100 18.8 .+-.
2.3 400.degree. C. 11.4 .+-. 1.0 600.degree. C. 6.2 .+-. 0.7 .sup.
800 C. 3.7 .+-. 0.2
[0078] Pyrolysis of electrospun SU-8-coated wires is a way to
generate a carbon fiber-coated surface for SPME. Just as in the
previous examples, in order to maintain the fiber structure through
pyrolysis, the samples must be cross-linked prior to pyrolysis.
[0079] The extraction capability of the electrospun-coated SPME
fibers was first examined with a mixture of benzene, toluene,
ethylbenzene, and o-xylene (BTEX). FIG. 12 shows the extraction
data obtained for the SU-8 electrospun fiber-coated wire. This
fiber-coated wire was not pyrolyzed but only conditioned in the
inlet at a temperature of 300.degree. C.
[0080] FIG. 13 shows the extraction time profile for the
400.degree. C. electrospun fiber-coated wire. The equilibration
time for BTEX with the 400.degree. C. fiber was approximately 30
minutes for benzene, toluene, ethylbenzene and o-xylene. FIG. 15
shows the extraction time profile for the 600.degree. C.
electrospun fiber-coated wire. The equilibration time for BTEX with
the 600.degree. C. fiber was approximately 60 minutes for benzene,
and 30 minutes for toluene, ethylbenzene, and o-xylene. FIG. 16
shows the extraction time profile for the 800.degree. C.
electrospun fiber-coated wire. The equilibration time for the
800.degree. C. fiber-coated wire was approximately, 30 minutes for
benzene, toluene, ethylbenzene, and o-xylene.
[0081] The extraction performance of the electrospun fiber-coated
wires was determined by comparing the fibers to the commercial
fibers of PDMS and PDMS/DVB. FIG. 11 shows a bar graph comparing
the extraction performance of each fiber for benzene. As shown in
FIG. 16, the 800.degree. C. electrospun fiber-coated wire was found
to have higher extraction efficiency for benzene as compared to the
other electrospun fiber-coated wires and the commercial fibers. It
is remarkable that not only did the 800.degree. C. have a higher
extraction efficiency compared to the commercial fiber but the
600.degree. C. fiber also had higher extraction efficiency.
[0082] The limit of detection and linear range for 600 and
800.degree. C. electrospun fiber-coated SPME wires were examined
for benzene and ethylbenzene. The detection limit of benzene
determined for both fibers were 0.3 ng/ml (600.degree. C.) and 0.4
ng/ml (800.degree. C.) and for ethylbenzene was determined to be
0.9 ng/ml (600.degree. C.) and 1 ng/ml (800.degree. C.). It is
interesting to point out that both of these fibers have a wide
linear range (0.05-40 .mu.g/ml) for a small coating thickness. The
large linear range can be attributed to the high surface area of
the nanofibers. The detection limits for both the 600 and
800.degree. C. electrospun fiber-coated fibers are comparable to
previously published SPME coatings but the linear range of these
coated wires was larger than previously reported SPME fibers.
[0083] For extracting phenol, 4-chlorophenol, and 4-nitrophenol,
the aqueous solution was acidified and saturated with NaCl as
above. The extraction time profiles for the SU-8, 400.degree. C.,
600.degree. C., and 800.degree. C. electrospun fiber-coated wires
were examined from 0 to 90 minutes for the above phenolic
compounds. The extraction profile for the SU-8 electrospun
fiber-coated SPME wire is shown in FIG. 16.
[0084] For the pyrolysis temperatures of 400, 600, and 800.degree.
C., the extraction time profile for the phenolic compounds are
shown below. The extraction time profile of the 400.degree. C. is
shown in FIG. 17. For the 600.degree. C. electrospun-coated SPME
fiber, the extraction time profile is shown in FIG. 18. The
extraction time profile for the 800.degree. C. SPME fiber is shown
in FIG. 19. The wide range of effective pyrolysis temperatures
demonstrate the high thermal stability of the electrospun
fibers.
[0085] The extraction performance of the four electrospun wires
compared to a commercially available polyacrylate (PA) SPME was
examined for the three phenolic compounds. A representative result,
that for 4-nitrophenol is shown in FIG. 20. Both the 600 and
800.degree. C. electrospun-coated SPME fibers displayed higher
extraction efficiencies than the PA fiber for phenol. For
4-chlorophenol, the 800.degree. C. fiber demonstrated superior
extraction performance as compared to the PA fiber; the extraction
efficiency of the 600.degree. C. was comparable to that of the PA
fiber. The 600 and 800.degree. C. electrospun coated-wires showed
greater performance than the PA fiber for 4-nitrophenol; the
efficiency of the SU-8 SPME fiber was comparable to that of the
commercial fiber.
[0086] Since it displayed adequate extraction efficiency for all of
the phenolic compounds, the SU-8 electrospun fiber-coated wire was
chosen to be analyzed for limit of detection and linear range for
phenol, 4-chlorophenol, and 4-nitrophenol. For the SU-8 fiber
coated SPME wire, the limit of detection was determined to be 3
ng/mL for phenol, 2 ng/mL for 4-chlorophenol, and 7 ng/mL for
4-nitrophenol. The SU-8 fiber-coated wire also showed a large
linear range for each of the compounds. It had a linear range of
0.004 .mu.g/mL to 4 .mu.g/mL for both phenol and 4-nitrophenol,
while 4-chlorophenol had a linear range from 0.01 .mu.g/mL to 10
.mu.g/mL. This demonstrates the ability of the fiber to be used for
a wide range of sample concentrations.
Electrospun Stationary Phase TLC
[0087] The electrospinning technique taught above also has
industrial applicability in the field of thin layer chromatography
(TLC). Electrospun fibers permit the thickness of the TLC
stationary phase to be reduced from the typical 150 to 250 .mu.m
range to a range of about 3 to about 30 The technique also allows a
significant reduction in the diameter, from the micro to the nano
scale, providing a method to tailor efficiency and speed of
separation.
[0088] In a first embodiment of a TLC technique that the fibers
facilitate, a TLC stationary phase was prepared by an
electrospinning technique as described above, but where a
polyacrylonitrile (PAN) was spun instead of SU8. The particular PAN
selected had a molecular weight of about 150,000. Anhydrous
dimethylformamide (DMF) was selected as the solvent for the PAN and
the electrospinning solution was 10% (wt/vol) PAN in DMF. An
aluminum foil was used as the collector substrate rather than a
silicon plate. It was demonstrated that a plate can be produced in
less than ten minutes with no sacrifice in quality or
reproducibility.
[0089] It will be understood from the ability to provide
appropriate fibers by electrospinning both a photo resist material
and a highly polar polymer such as a polyacrylonitrile, that the
technique taught herein may be applied to vary the nature of a TLC
stationary phase to include a stationary phase with polar moieties,
nonpolar interactions, hydrogen bonding interactions, adsorptive
behavior, high porosities, etc. Further, the work has demonstrated
that the electrospun fibers can be successfully pyrolyzed to
provide carbon nanofibers for TLC stationary phase.
[0090] Materials that were selected for TLC using the PAN
electrospun plate include the steroids cholesterol, androsterone
and cortisone. For the testing, the mobile phases examined were a
normal phase, consisting of mixtures of petroleum ether, with
acetone, ranging from 0 to 100% acetone and a reversed phase,
consisting of acetone and water mixtures ranging from 0 to 100%
acetone.
[0091] Retardation factors for separation of the steroid analytes
in the normal phase system using the electrospun PAN stationary
phase are reported below. The results in the range of 30 to 50%
acetone are not shown, as there was extreme horizontal broadening
within that range. As a point of comparison, the same steroids were
separated by the same normal phase mixtures in 10% acetone
increments from 0 to 100% acetone.
[0092] Retardation factors for separation of the steroid analytes
in the reversed phase system of water in acetone, using the
electrospun PAN stationary phase are shown below. Unlike the normal
phase experiment, the electrospun PAN plate was useful across the
entire range of from 0 to 100% water, with experiments conducted at
10% increments of water in acetone. The same steroids were
separated by the same reversed phase mixtures in 10% acetone
increments from 0 to 100% acetone. In the reversed phase
experiment, it is notable that cholesterol, the least polar
compound, is the only one completely retained on the electrospun
PAN plate at any point in the experiment, in contrast to the
performance of the steroid analytes on the conventional TLC plate.
It is further notable that cortisone, as the most polar analyte, is
retained almost uniformly through the entire mobile phase range on
the electrospun PAN plate, while the retardation factor for
cortisone declines in a similar fashion to the other analytes on
the conventional TLC plate.
TABLE-US-00002 Solvent Compound Distance distance Ave. Rf Ave. w
Ave. k' Ave. N Cholesterol 0 32 0 3.4 0 0 Andros- 11.6 32 0.36 2.6
1.9 625 terone Cortisone 22.4 32 0.7 3.6 0.44 1000
[0093] A further experimental confirmation of the efficacy of the
electrospun PAN TLC plate, compared to conventional TLC, was
provided by selecting a set of 7 laser dyes from Exiton, Inc. as
the test analytes. These analytes are all fluorescent in the
visible range when excited by UV radiation. The seven selected dyes
are Sulforhodamine 640, (S640) Pyromethene 597 (P597), Rhodamine
610 perchlorate (R610P), Rhodamine 610 chloride (R610C), Rhodamine
590 chloride (R590C), Rhodamine 101 (R101) and Kiton Red (KR). From
the data obtained, the following table of retardation factors was
calculated:
TABLE-US-00003 Ave Rf Silica Ave Rf E- % RSD Rf % RSD Rf E-
Compound plate spun plate Silica plate spun plate S640 0.66 0.69
4.7 6.8 P597 0.81 0.67 19 9.7 R610P 0.16 0.59 18 13 R610C 0.16 0.6
13 23 R590C 0.50 0.42 35 24 R101 0.08 0.64 23 23 KR 0.63 0.67 5.5
15
[0094] For the same compounds, the experiments provide a comparison
of efficiencies between a conventional silica plate and the
electrospun embodiment:
TABLE-US-00004 Ave N Silica Ave N E-spun % RSD N % RSD N E-
Compound plate plate Silica plate spun plate S640 9400 2000 79 58
P597 1300 2400 42 110 R610P 51 390 57 86 R610C 85 970 66 110 R590C
18 120 59 190 R101 29 970 64 99 KR 1400 1500 52 90
[0095] From the foregoing dye data, a few notable points are that
the retardation factors and precisions obtained using the
electrospun plates compare favorably to the conventional silica
plates, that the electrospun plates exhibit higher efficiencies,
although at a larger uncertainty level, and that complete
separations are accomplished in 15 of 21 mixtures by the
electrospun compared to 19 of 21 complete separations using the
silica plate. The electrospun plate was able to completely separate
S640 from KR, while a silica plate could not, and neither TLC plate
could completely separate R610P from R610C.
Carbon Nanofiber Stationary Phase Thin Layer Chromatography
[0096] While still in widespread use today, TLC has drawbacks.
Recent advances in TLC technology such as High Performance TLC
(HPTLC) and Ultrathin TLC (UTLC) have helped to address some of
these weaknesses. HPTLC allows for TLC stationary phases with
smaller dimensions, allowing TLC to be used for quantitative as
well as qualitative analysis. UTLC is a more recently developed
method and has been found to further improve sensitivity and to
reduce both analysis time and amount of consumable required for
TLC.
[0097] Commercially available TLC plates are typically composed of
a stationary phase with a thickness that ranges from 100-400 .mu.m
whereas UTLC plates created using monolithic silica gel or other
sorbent materials have layers in the 5-30 .mu.m range. The main
drawback of UTLC is the low sample capacity. Many materials have
been studied for use in TLC, including activated carbon which
encountered issues with retention mechanisms of the analytes, the
need for more exotic binders to adhere the new mixtures to the
plates, and when coupled with silica gel, it was found that the
percentage of activated carbon could not exceed 16%.
[0098] Like activated carbon, glassy carbon (also termed
graphitized carbon) has potential as a sorbent for chemical
separation. Moreover, like graphite, glassy carbon is a
hydrophobic, yet highly polarizable solid. As such, glassy carbon
can retain very polar analytes, such as cations and anions, which
conventional C18 sorbents cannot, making glassy carbon especially
desirable as a reversed phase sorbent for chromatography.
[0099] Conventional TLC systems, including those that employed
activated carbon as a sorbent generally require binders or support
materials to adhere the stationary phase to the plate. However,
electrospinning polymers directly to the support plate and
subsequent conversion of the polymer nanofibers into carbon
nanofibers, reproducibly delivers a suitable medium for UTLC
separation and analysis without binders.
[0100] Silicon wafers or stainless steel plates were cut into
rectangular pieces; the substrate was placed on a ring stand and
grounded. Each TLC plate was electrospun with SU-8 2100 solutions
for either 15 or 20 minutes. After the electrospinning was
finished, the plates were pyrolized using substantially the same
methods discussed above.
[0101] The plates were then spotted with .about.50 nL of the sample
using a 250 .mu.m internal diameter glass capillary and developed
in 5 mL of mobile phase. The mobile phase was allowed to
equilibrate with the vapor phase for 10 minutes prior to analysis.
The mobile phase used for the laser dye studies was 2-propanol
(Sigma-Aldrich, Atlanta, Ga.). 20 mM borate buffer formed in 90:10
water to acetonitrile was used as the mobile phase for amino acid
analysis.
[0102] In order to overcome problems of carbon stationary phase
(dark color) and visualize the analytes on the carbon stationary
phase, a small amount of water or mobile phase was placed on top of
the plate while it was placed horizontally and viewed under UV
radiation.
[0103] The test analytes were six laser dyes acquired from Exciton
Inc. (Dayton, Ohio) and were examined in two groups containing
structurally similar compounds, Group A.: rhodamine 610 perchlorate
(CAS: 23857-51-4), rhodamine 610 chloride (CAS: 81-88-9) and kiton
red (CAS: 3520-42-1); and Group B.: sulforhodamine 640 (CAS:
60311-02-6), pyrromethene 597 (CAS: 137829-79-9), and rhodamine 101
(CAS: 72102-91-1). DL-phenylalanine, L-lysine and L-threonine were
available commercially. An unmodified standard silica phase with
particle sizes of 5-17 BET surface area of .about.500 m.sup.2/g,
mean pore size of 60 .ANG., 0.75 mL/g pore volume, and layer
thickness of 200 .mu.m was used.
[0104] Fibers resulting from three different pyrolysis
temperatures: 600, 800 and 1000.degree. C., were examined. Fiber
diameters and mat thicknesses for each pyrolysis temperature are
shown in Table 1. The average fiber size of the glassy carbon
fibers decreases with pyrolysis temperature which is consistent
with typical observations of the evolution of low molecular weight
compounds, condensation and shrinkage that occurs when converting
polymers into carbon.
[0105] Raman spectroscopy was used to characterize the glassy
carbon nanofiber stationary phase resulting from different
pyrolysis temperatures. Characteristic Raman spectra of electrospun
SU-8 samples pyrolyzed at 600.degree. C. (A) and 1000.degree. C.
(B) are shown in FIG. 20. They show the classic D (.about.1350
cm.sup.-1) and G (.about.1580-1600 cm.sup.-1) bands associated with
glassy carbon. There is an obvious shift in the ratio of the peak
areas of the D and G peaks as the pyrolysis temperature
increases
[0106] In order to analyze the stationary phase thickness, a TLC
device was fabricated and then the substrate was cleaved and placed
on its side on an SEM stage in order to view the stationary phase
in an "edge on" manner. A representative image is shown in FIG.
19.
[0107] Each plate was used for at least 35 trials before
performance or physical structure diminished. After the plates were
no longer usable, the carbon nanofiber stationary phase was simply
removed by scrubbing the substrate and then cleaning with
acetone.
[0108] One of the advantages of ultrathin-layer chromatography is
the decrease in analysis time due to increased mobile phase
velocity. FIG. 21 shows the migration distance of the mobile phase
versus the square root of the migration time. This plot compares
the relative mobile phase velocities in each of the carbon devices
as well as the PAN UTLC device and a conventional silica gel
plate.
[0109] When comparing the glassy carbon devices, it is clear that
the largest observed fiber diameters demonstrate the fastest mobile
phase velocities and this transport velocity decreases as a
function of fiber diameter. The largest average fiber diameters are
seen in the 600.degree. C. device followed by the 800.degree. C.
and 1000.degree. C. devices. The electrospun PAN UTLC device shows
a mobile phase velocity very similar to the velocity observed for
the 600.degree. C. carbon device and the average fiber diameters
for the two devices were of similar magnitude. The commercial
silica gel TLC plate shows a mobile phase velocity similar to the
1000.degree. C. carbon device.
[0110] In this study, electrospun glassy carbon UTLC devices are
compared with electrospun PAN UTLC devices for the analysis of
laser dyes in FIG. 22. The highest efficiency varied based on
different analytes but was typically obtained using the 800 or
1000.degree. C. carbon plate. It is noteworthy that for each dye,
at least one of the electrospun glassy carbon UTLC devices
demonstrates higher efficiency than the PAN UTLC device which was
previously shown to be a highly efficient technique.
[0111] With the exception of one compound, rhodamine 101, the
efficiency increased as a function of processing temperature. In
all cases, the efficiency increased between the 600.degree. C.
plate and the 800.degree. C. plate. All analyses using the
electrospun glassy carbon stationary phase showed greater
efficiency than the PAN UTLC phase which was also substantially
higher than silica TLC devices commercially available.
[0112] The retardation factors for replicate studies with, n=7, for
each laser dye are shown in FIG. 23. The plots illustrate that the
general trend for both groups of dyes is a decrease in the Rf value
with increasing pyrolysis temperature. This means that the laser
dyes are more strongly retained as the pyrolysis temperature
increases. The decrease in retardation factor of the highly
aromatic laser dye compounds is expected due to stronger .pi.-.pi.
and dispersive interactions at higher processing temperatures. The
data are supported by previous studies of glassy carbon that show
polarizability becomes an important factor at higher
temperatures.
[0113] The electrospun carbon UTLC technique showed a much higher
resolution for all components of the laser dye mixture when
compared to the electrospun PAN UTLC method.
[0114] Fluorescein isothionate (FITC) labeled amino acid solutions
were prepared according to known procedures. The separations were
carried out using a mobile phase composed of a 90:10 water to
acetonitrile borate buffer solution with a pH of 9.5.
[0115] FIG. 24 presents the efficiency obtained using the glassy
carbon UTLC devices in comparison to work by Khan et al. The 600
and 800.degree. C. carbon devices show a vast increase in
efficiency when compared to the mixed cellulose phase. It is also
worthy to note that highly efficient analysis were obtained using
the glassy carbon UTLC technique using development times of 5
minutes and development distances of 3.5 cm while the comparison
work used a development distance of 15 cm. The 1000.degree. C.
plate showed extreme band broadening and was clearly not optimal
for essential amino acid analysis at these concentrations.
[0116] In order to examine the cause of the poor performance of the
1000.degree. C. plates, the sample capacity of each of the
electrospun glassy carbon UTLC devices was investigated. It was
found that the 600 and 800.degree. C. plates performed
satisfactorily in the concentration range of 10.sup.-7 to 10.sup.-5
M while the 1000.degree. C. device was overloaded at concentrations
higher than 10.sup.-6 M.
[0117] FIG. 25 shows the Rf values for each of the amino acids. The
plot demonstrates the same trend as in the analysis of the laser
dyes. The retardation factor of each amino acid decreases with
increasing pyrolysis temperature. It is also interesting to note
that the order of elution, Phe-Thr-Lys, changes to Thr-Phe-Lys
between 600 and 800.degree. C. This change in selectivity can be
attributed to the increased sp.sup.2-character of the stationary
phase at 800 in comparison to the 600.degree. C. plate. This is
born out by the Raman spectra of these samples. These results
illustrate the ability of the glassy carbon nanofibrous stationary
phase to have tunable selectivity depending upon processing
temperature.
[0118] FIG. 26 demonstrates the separations obtained with the
electrospun glassy carbon UTLC devices and comparison to previously
published literature values. The 800.degree. C. devices showed the
most desirable chromatographic performance for high efficiency
separations of the ternary essential amino acid mixture. The
electrospun carbon UTLC provides higher efficiency and resolution
when compared to previous analyses. The results suggest that the
mixture could be completely resolved at even shorter analysis times
and development distances.
Additional Experimentation
[0119] Glassy carbon (GC) or porous graphitic carbon (PGC) are
widely-used sorbents in chromatography. Due to its tunable
retention behavior and both chemical and physical stability, glassy
carbon has been used for HPLC, supercritical fluid chromatography
(SFC), gas chromatography, as well as, solid phase extraction and
microextraction (SPE and SPME). The most commonly-used type of
glassy carbon is PGC, which is a very expensive sorbent that is
fabricated by coating porous silica particles with a glassy carbon
precursor, pyrolyzing the particles, then dissolving the silica
particles leaving behind a porous glassy carbon material.
[0120] The glassy carbon nanofibers used herein have the same
molecular structure as PGC, but are macroscopically non-porous. The
electrospun glassy carbon nanofibers, like PGC, are amorphous and
only demonstrate repeatable structure at nanoscopic levels. Glassy
carbon surfaces are sheets of hexagonally oriented
sp.sup.2-hybridized carbon atoms. Nanoscopically, glassy carbon is
composed of thin ribbons, with widths less than 10 nm. Like
graphite, glassy carbon has two different types of carbon atoms
those located on the basal or edge-planes. Both carbons are
sp.sup.2-hybridized. The extent of delocalization of the
sp.sup.2-hybridization can be controlled by processing temperature,
where higher temperatures yield enhanced delocalization.
[0121] Glassy carbon has not been used in thin-layer chromatography
due in part to the detection problem that results from its dark
color and ability to quench fluorescence. Past TLC work using
activated carbon was performed by mixing with silica gel. This
mixed phase was not ideal for TLC as the activated carbon to silica
gel ratio strongly affected the retention mechanism and dispersion
of the analytes. This study also showed that the maximum amount of
carbon that could be incorporated into the stationary phase was
16%. So although activated carbon has previously been incorporated
into a TLC method, to the best of our knowledge, this study is the
first to employ a pure glassy carbon stationary phase for TLC. This
work introduces a technique for the creation and use of an
electrospun nanofibrous glassy carbon stationary phase that does
not require support material or binders for use in reversed-phase
ultrathin layer chromatography.
[0122] The polymer used for the electrospinning experiments was
SU-8 2100 negative photoresist (an epoxide polymer). SU-8 2100 was
purchased from MicroChem Corp. (Newton, Mass.). The solvent used to
prepare the dilute SU-8 2100 solutions was cyclopentanone
(Sigma-Aldrich, Atlanta, Ga.). The test analytes were six laser
dyes acquired from Exciton Inc. (Dayton, Ohio) and were examined in
two groups containing structurally similar compounds; Group A:
rhodamine 610 perchlorate (CAS: 23857-5194 4), rhodamine 610
chloride (CAS: 81-88-9) and kiton red (CAS: 3520-42-1); and Group
B: sulforhodamine 640 (CAS: 60311-02-6), pyrromethene 597 (CAS:
137829-79-9), and rhodamine 101 (CAS: 72102-91-1). DL-phenylalanine
from Eastman Chemical (Kingsport, Tenn.), L-lysine from
Sigma-Aldrich and L-threonine from Calbiochem (Los Angeles, Calif.)
were used. Electrospun polyacrylonitrile (PAN) UTLC plates were
fabricated in house as described previously. Two
commercially-available thin layer chromatography plates were
purchased from Macherey-Nagel: a cyano functionalized silica
(particle diameters of 2-10 .mu.m, .about.500 m.sup.2/g BET surface
area, 60 .ANG., 0.75 mL/g pore volume, and layer thickness of 200
.mu.m on aluminum backing: (catalog number 818184) and an
unmodified silica phase (catalog number 818133) with diameters of
5-17 .mu.m and other physical attributes the same as the cyano
modified phase.
[0123] Electrospinning is a simple and inexpensive method of
producing nanofibers 108 composed of a polymeric material. All
electrospinning experiments were conducted in a plexiglass
enclosure which was used to control the temperature and humidity of
the environment, as well as, serving to isolate the nanomaterials
inside a closed system as a safety precaution. According to OSHA
standards 1910.132 and 1910.134, air purifying respirators with
cartridge filters having a sufficient assigned protection factor
(APF) should be used when working with engineered nanomaterials
when they are produced in the open lab. By maintaining the spinning
process inside the plastic box no exposure is possible except when
the process is completed and the box is opened. The electrospun
fibers have one dimension which is not nano-sized. The fibers have
diameters of .about.400 nm, but they have macroscopic lengths that
vary from the centimeter to meter range. Thus, a respirator is used
purely as a precaution and the plastic enclosure is used to further
reduce any potential risks. The optimized parameters for
electrospinning SU-8 were previously examined in this group and
found to be 75% (vol/vol) SU-8 in cyclopentanone, 0.02 mL/min flow
rate, 9 kV, and 10 cm tip to collector distance. Because of the
photo sensitive nature of SU-8 2100, all electrospinning
experiments were performed in a dark room under yellow light. Due
to differences in mass loss and shrinkage caused by varying
processing temperatures, the electrospinning time was optimized for
each device. In order to obtain a mat of fibers with a consistent
thickness, devices pyrolyzed at 600.degree. C. were spun for 15
minutes while devices processed to 800.degree. C. and 1000.degree.
C. were electrospun for 20 minutes each. These conditions result in
an electrospun glassy carbon nanofibrous mat with a thickness of
.about.15 .mu.m after heat processing.
[0124] Stainless steel plates were cut into rectangular pieces
approximately 2 cm.times.6 cm and cleaned with acetone prior to
electrospinning. A plate was placed on a ring stand and grounded
using an electrode from the high voltage power supply. After the
electrospinning was finished, the SU-8 was crosslinked by exposing
the fibers to ultraviolet light for 5 minutes. Previous work
illustrates that electrospun SU-8 fibers do not retain their
structure during pyrolysis unless this photo-polymerization step is
taken. After the electrospun fibers were exposed to the UV
radiation, they were placed directly into a tube furnace
(Lindberg/Blue Waltham, Mass.) equipped with a PID (proportional,
integral, derivative) temperature controller and heated to the
specified pyrolysis temperature at a ramp rate of 2.degree. C. per
minute under a forming gas (N2/H2) environment. The plates were
held at the pyrolysis temperature for at least 6 hours, then the
oven was turned off and the samples were allowed to cool to room
temperature. FIG. 27A-C illustrates the benchtop view of the
electrospun glassy carbon UTLC plate.
[0125] The electrospun carbon UTLC plates were used without further
treatment after the pyrolysis was complete. The plates were spotted
with .about.50 nL of the sample using a 250 .mu.m internal diameter
glass capillary and developed in 5 mL of mobile phase. The mobile
phase was placed into a sealed cylindrical glass jar with an
internal volume of 200 mL. The mobile phase depth in the
development chamber was 3 mm. The chamber was sealed and allowed to
equilibrate with the vapor phase for 10 minutes prior to analysis.
This resulted in saturated conditions for the experiments. Mobile
phase conditions were optimized for maximum resolution for both
sample sets. The optimized mobile phase used for the laser dye
studies was 2-propanol (Sigma-Aldrich, Atlanta, Ga.) while 20 mM
borate buffer formed in 90:10 (volume:volume) water to acetonitrile
was used for amino acid analysis. The development of a plate took
approximately 5 minutes for each trial. After development, the
plates were removed from contact with the mobile phase and dried at
room temperature.
[0126] A digital documentation system obtained from Spectroline was
used for all UTLC visualization: CC-81 cabinet with ENF-280C 365
nm/254 nm, 8 Watt UV Lamp; GL-1301 universal camera adapter with 58
mm adapter ring along with a Canon A650IS 12.1 MP digital camera to
collect the images. Data was obtained by obtaining a photograph of
each separation using a digital photograph, analyzing the image and
obtaining raw data using TLC Analyzer, then using PeakFit to
analyze the chromatographic parameters of the separation. Analyses
of digital images were performed using TLC Analyzer software
available at:
http://www.sciencebuddies.org/science-research-papers/tlc.sub.--analyzer.-
shtml. A digital image is made up of many rows and columns of
pixels. Thus, a digital image is essentially a matrix of numbers,
and TLC Analyzer analyzes images by panning across the matrix and
combining moving averages to create a chromatogram. Representative
digital photographs and/or the resulting chromatograms are shown in
FIGS. 30 and 32.
[0127] The lack of commercially-available electrospun and/or carbon
TLC devices present a challenge for comparison of our results. In a
recent comprehensive review, only two methods for preparing UTLC
devices are reported, and neither one is commercially-available. As
a result of this, the electrospun carbon UTLC devices were compared
to polymer UTLC devices using electrospun fibers, as well as, the
two commercially-available, silica based sorbents from
Macherey-Nagel.
[0128] All nanoscopic images of the electrospun fibers were
obtained using a Hitachi S-4300 (Hitachi High Technologies,
America, Inc., Pleasanton, Calif., USA) scanning electron
microscope (SEM). The electrospun SU-8 fibers were pyrolyzed using
a Lindberg Blue (Waltham, Mass.) tube furnace. A Sunray 400SM UV
flood light (Uvitron International West Springfield, Mass.) was
used to crosslink the SU-8 fibers before pyrolysis. Raman spectra
were obtained using a Renishaw InVia Raman Microscope
(Gloucestershire, UK) with Argon laser excitation at 514.5 nm.
[0129] Electrospun carbon UTLC plates processed to three different
pyrolysis temperatures: 600, 800 and 1000.degree. C., were
examined. A representative SEM image of the fibers is shown in FIG.
1B. Fiber diameters and mat thicknesses for each pyrolysis
temperature are shown in Table 1.
TABLE-US-00005 TABLE 1 Description of the all devices used for TLC
studies. (*Particle Diameter) Mat Thickness Avg. Fiber Device
(.mu.m) Diameter 600.degree. C. 16 .+-. 1.4 330 .+-. 70 nm
800.degree. C. 14 .+-. 1.0 300 .+-. 70 nm 1000.degree. C. 13 .+-.
1.5 220 .+-. 70 nm PAN UTLC 24 .+-. 1.8 395 .+-. 55 nm Silica Gel
200 2-10 .mu.m*
The average diameter of the glassy carbon fibers decreases with
pyrolysis temperature which is consistent with typical observations
of the evolution of low molecular weight compounds, condensation
and shrinkage that occurs as the carbon is further graphitized.
[0130] In order to analyze the stationary phase mat thickness, a
TLC device was fabricated and then the substrate was cleaved and
placed on its side on an SEM stage in order to view the stationary
phase in an "side-view" manner. A representative image is shown in
FIG. 27C.
[0131] To confirm the glassy carbon structure of the SU-8
nanofibers after pyrolysis, Raman spectroscopy was performed. A
detailed description of the analyses and the resulting spectra can
be found in the Supplemental Information section. The findings were
consistent with other reports of glassy carbon-like materials. From
the Raman spectra, the average microcrystallite size, La, (shown in
Table 2) of the glassy carbon was calculated using the Tuinstra and
Koenig equation and was found to decrease with increasing
processing temperature. La was determined to be between 1 and 3 nm
for each of the electrospun glassy carbon UTLC plates, which is
typical of glassy carbon at processing temperatures between
600.degree. C. and 1000.degree. C.
TABLE-US-00006 TABLE 2 Comparison of effective pore radii,
microcrystallite size, and .kappa. values for each of the TLC
devices examined. TLC Device Effective Pore Radius, R L.sub..alpha.
.kappa. (cm2/sec) Silica Gel 345 .+-. 20 nm N.A. 0.019 PAN UTLC 515
.+-. 10 nm N.A. 0.028 600.degree. C. 475 .+-. 35 nm 2.99 nm 0.026
800.degree. C. 400 .+-. 15 nm 2.39 nm 0.022 1000.degree. C. 380
.+-. 25 nm 1.47 nm 0.020 Heptane was the mobile phase for the
calculation of R. N.A.--not applicable.
When glassy carbon is heated, the sp.sup.2-hybridization of the
carbon lattice increases which causes enhanced delocalization of
the .pi. electron cloud. Accordingly, enhanced .pi.-.pi.
interactions with analytes also occurs with increased processing
temperature. The fabrication of electrospun glassy carbon UTLC
plates in this manner allows for the UTLC devices to have tunable
retention dependent upon processing temperature.
[0132] The electrospun carbon UTLC devices proved to be chemically
and mechanically robust over a large number of experiments. Each
plate was used for at least 35 trials before performance or
physical structure diminished. After the plates were no longer
usable, the carbon nanofiber stationary phase was simply removed by
scrubbing the substrate and then cleaned with acetone. A new device
for UTLC could then be created by electrospinning onto the reusable
substrate.
[0133] The Lucas-Washburn equation (equation 1) is used to describe
the capillary flow through porous media:
Z f 2 = .gamma. Rt cos .theta. 2 .eta. ( 1 ) ##EQU00001##
[0134] In this equation, Zf is the distance travelled by the
solvent front, t is the corresponding time, .gamma. and .eta. are
the surface tension and viscosity of the mobile phase, .theta. is
the contact angle of the mobile phase with the stationary phase,
and R is the effective pore radius.
[0135] In order for the Lucas-Washburn equation to accurately
describe a system, the following must be true: laminar flow
conditions must prevail (Reynolds number<1200); an adsorbed film
of penetrating liquid must exist (i.e. must expose to vapors of
solvent and reach equilibrium before use); inertial influences must
be negligible, and the contact angle of mobile phase on the surface
must be 0. Equilibrium conditions apply in this study and thus all
of these conditions are satisfied, as they are in most TLC
experiments. To examine our system, plots of Zf.sup.2 versus t are
shown in FIGS. 28 and 29. FIG. 28 is a comparison of a 600.degree.
C. glassy carbon UTLC device with an electrospun PAN UTLC device as
well as a commercially-available silica TLC plate. FIG. 29 compares
the 600.degree. C., 800.degree. C. and 1000.degree. C. electrospun
glassy carbon UTLC devices. The linear fits show agreement with the
Lucas-Washburn equation (r.sup.2>0.98 for each).
[0136] When comparing the mobile phase migration using electrospun
glassy carbon UTLC devices, FIG. 29, it is clear that the largest
observed fiber diameters demonstrate the fastest mobile phase
velocities and the transport velocity decreases as a function of
fiber diameter (600.degree. C.>800.degree. C.>1000.degree.
C.). For example, the 600.degree. C. device migrates .about.15%
farther than the 1000.degree. C. device for the same development
time. The electrospun PAN UTLC devices show a mobile phase velocity
very similar to the velocity observed for the 600.degree. C. carbon
device and the average fiber diameters for the two devices were of
similar magnitude. The commercial silica gel TLC plate shows a
mobile phase velocity similar to the 1000.degree. C. carbon
device.
[0137] Kalogianni and co-workers previously used the Lucas-Washburn
equation to measure the effective pore radii of porous media using
nonpolar solvents which had 0.degree. contact angles with the
measured surfaces. Each mobile phase used for separations, as well
as heptanes, showed 0.degree. contact angles on the glassy carbon
surfaces. Therefore, heptanes were used as the mobile phase to
calculate effective pore radii for the three carbon devices, the
electrospun PAN UTLC device, and a commercially available silica
gel TLC plate. Using the Lucas-Washburn equation the effective pore
radii, R were calculated (Table 2). Interestingly, the estimated
effective pore radii, R, for the glassy carbon devices were
approximately two to three times larger than the observed fiber
radii that were measured using SEM while the calculated R for the
commercially-available silica phase is .about.1/3 of the particle
diameter. The pore structures for the nanofibers and the
commercially-available phases are clearly very different.
[0138] The velocity constant, .kappa., is often used to describe
mobile phase transport in TLC. .kappa. is a constant that replaces
the term .gamma.R/2.eta. in the Lucas-Washburn equation and also
takes into account the average particle diameter, pore size
distribution and permeability of the layer for typical precoated
TLC layers. The velocity constant values were determined for each
of the electrospun glassy carbon UTLC devices, an electrospun PAN
UTLC device and a commercially-available silica gel TLC plate. As
can be seen in Table 2, the values experimentally obtained for each
of the commercially available TLC devices are similar to those
previously reported for silica TLC plates using similar organic
solvents and are on the order of 10.sup.-2 cm.sup.2/sec. However,
all of the electrospun UTLC devices (PAN, as well as each glassy
carbon device) demonstrate larger .kappa. values than the
commercially available silica gel plate meaning that faster mobile
transport occurs in the electrospun UTLC devices. When a direct
comparison of migration distance versus time is made, the
600.degree. C. device is found to take approximately half of the
time (3.6 min for electrospun UTLC) required to reach a development
distance of 3.5 cm when compared to commercially-available silica
plates (7 min).
[0139] The procedure for applying samples to each TLC plate is
described above in section 2.4. The average starting spot width for
the electrospun glassy carbon UTLC devices were 0.25-0.5 mm for the
laser dyes and 0.5 mm for amino acids. Using the same sample
application process on the silica leads to starting spot widths of
1 mm and 0.5 mm for the laser dyes and amino acids, respectively.
Typically, resulting spot widths on the glassy carbon UTLC devices
at the end of the development was 0.5 to 1 mm, with the largest
average spot size being lysine at 2.85 mm on the 800.degree. C.
UTLC device. Therefore, the developed spot widths on the glassy
carbon plate are similar to the initial sample spot widths.
[0140] Carbon is a fluorescence quencher. In order to overcome this
problem and to visualize the analytes on the carbon stationary
phase, a small amount of water or acetone was sprayed on top of the
plate using a TLC reagent nebulizer (Kimble-Chase Vineland, N.J.)
while the plate was horizontal and viewed under UV radiation.
Addition of liquid to the dry plate causes desorption of the
analytes from the glassy carbon stationary phase and their
fluorescence can be observed. In order to check the validity of
this visualization process, digital images were captured at 2
second intervals after the introduction of the liquid to the plate.
A laser dye, sulforhodamine 640, was spotted on an electrospun
carbon UTLC plate and then acetone was sprayed on the plate as
digital images were captured every two seconds. The images were
then analyzed using TLC Analyzer software and were found to have
similar spot sizes (i.e. less than 5% increase in spot size) for up
to 10 seconds after the visualization liquid was sprayed onto the
plate, this was ample time for the digital camera to capture the
analyte spots prior to loss of spot structure.
[0141] Laser dyes from Exciton Inc. were used as test analytes for
the carbon TLC. Analysis of laser dyes is a common practice in
development of new TLC techniques. The dyes were an ideal test set
for the electrospun glassy carbon UTLC technique due to their
ability to fluoresce with UV excitation, their high quantum yields,
and their highly aromatic character. The dyes were initially
examined individually as a proof of concept to demonstrate that
analyte migration and differential retention were possible on the
electrospun glassy carbon UTLC devices. Plate number, N, and
retardation factor data presented here describes these
experiments.
[0142] The retardation factor was calculated using equation 2 where
ZS is the distance 318 travelled by the analyte spot and ZF is the
distance travelled by the solvent front:
R f = Z S Z F ( 2 ) ##EQU00002##
[0143] The retardation factors and standard deviation for replicate
studies with, n=7, for each laser dye are shown in Table 3.
TABLE-US-00007 TABLE 3 Retardation factors (R.sub.f) of laser dyes
on each of the electrospun carbon UTLC devices. TLC R .sub.f Device
Rh 610 P Rh 610 Cl KR S 640 P 597 Rh 101 600.degree. C. 0.56 .+-.
0.04 0.65 .+-. 0.10 0.44 .+-. 0.04 0.27 .+-. 0.04 0.83 .+-. 0.07
0.35 .+-. 0.04 800.degree. C. 0.32 .+-. 0.03 0.33 .+-. 0.03 0.29
.+-. 0.04 0.25 .+-. 0.07 0.70 .+-. 0.30 0.30 .+-. 0.05 1000.degree.
C. 0.38 .+-. 0.05 0.28 .+-. 0.05 0.37 .+-. 0.03 0.20 .+-. 0.05 0.60
.+-. 0.25 0.25 .+-. 0.04
The data illustrate that the general trend for both groups of dyes
is a decrease in the Rf value with increasing pyrolysis
temperature. This means that the laser dyes are more strongly
retained as the pyrolysis temperature increases. For example, the
retention of rhodamine 610 chloride increases by .about.50% between
processing temperatures of 600 and 800.degree. C. The decrease in
retardation factor (increase in retention) of the highly aromatic
laser dye compounds is expected due to stronger .pi.-.pi.
interactions that occur with increased processing temperature. Not
surprisingly the least retained (Rf=0.83) compound is pyrromethene
597 which is composed of a three ring system compared to all other
compounds which have at least four. The most strongly retained
compounds are sulforhodamine 640 and rhodamine 101 which are
structurally similar. These two compounds are composed of 8
six-figured rings and thus interact strongly with the glassy carbon
stationary phase via .pi.-.pi. and dispersive interactions.
[0144] In TLC, plate number, N, is directly proportional to the
square of the ratio of analyte (solute) migration distance, ZS, to
spot width, w.
N = 16 ( Z S w ) 2 ( 3 ) ##EQU00003##
Typically, conventional silica gel TLC and HPTLC devices developed
using capillary flow have N.about.250-5000. Poole predicted that
plate numbers approaching 100,000 are possible using forced-flow
HPTLC with 3 .mu.m diameter particles and development times of 25
minutes. However, this would also require large pressure drops
across the device. Electrospun UTLC is now providing high
efficiency without the need of forced-flow HPTLC.
[0145] A previous direct comparison of commercially-available
silica plates and electrospun polyacrylonitrile (PAN) UTLC plates
with similar stationary phase chemical surface structure showed a
vast improvement (1-3 orders of magnitude improvement in half the
analysis time) in the efficiency of analysis of laser dyes and
steroidal compounds. This increase in plate number was due in large
part to the decrease in spot width (typically a factor of 5)
demonstrated using the technique. The electrospun glassy carbon
UTLC devices in this study show an even greater decrease in
dispersion and therefore, spot width. A digital photograph of the
separation of sulforhodamine 640, rhodamine 610 perchlorate and
pyrromethene 597 was analyzed and the resulting chromatogram is
shown in FIG. 30.
[0146] The plate numbers of electrospun glassy carbon UTLC devices
processed at 600, 800 and 1000.degree. C. are compared with those
of electrospun PAN UTLC devices for the analysis of laser dyes in
FIG. 31. For a given analyte, the processing temperature that gave
the highest efficiency varied but was always obtained using either
800 or 1000.degree. C. processed electrospun glassy carbon
plates.
[0147] The resolution of a mixture of sulforhodamine 640, rhodamine
610 perchlorate, and pyrromethene 597 on the carbon devices and the
electrospun PAN UTLC device, is shown in FIG. 31. The mixture was
fully separated on each of the three carbon TLC plates. There is an
increase in resolution as the processing temperature increases due
to increased selectivity and higher efficiency separations. The
electrospun carbon UTLC technique proved to be a much higher
resolution method for all components of the laser dye mixture when
compared to commercially-available TLC plates. To summarize, the
electrospun glassy carbon UTLC plates show minimal band broadening,
resulting efficiencies in the range of 10,000 and also prove to be
a very high resolution technique for the analysis of laser
dyes.
[0148] In order to study essential amino acids on the electrospun
glassy carbon UTLC devices, lysine, threonine, and phenylalanine
were labeled with fluorescein isothionate (FITC), using a
previously reported method. FITC is commonly-used to fluorescently
tag biomolecules such as amino acids and proteins. Separations were
carried out using an optimized mobile phase composed of a 90:10
(volume:volume) water to acetonitrile borate buffer solution with a
pH of 9.5. FIG. 33 is a chromatogram of the separation of lysine,
threonine and phenylalanine on a 600.degree. C. electrospun glassy
carbon UTLC device.
[0149] Table 4 shows the Rf values for each of the amino acids.
TABLE-US-00008 TABLE 4 Retardation factors (R.sub.f) of amino acids
on each of the electrospun carbon UTLC devices. R.sub.f TLC Device
Lys Thr Phe 600.degree. C. 0.64 .+-. 0.04 0.91 .+-. 0.04 0.79 .+-.
0.06 800.degree. C. 0.59 .+-. 0.06 0.72 .+-. 0.22 0.79 .+-. 0.23
1000.degree. C. 0.56 .+-. 0.04 0.50 .+-. 0.22 0.51 .+-. 0.24
The data demonstrate the same trend is as in the analysis of laser
dyes. The retardation factor of each amino acid decreases with
increasing pyrolysis temperature. Not surprisingly, lysine is the
most strongly retained of the sample set. Under the separation
conditions, lysine is the only analyte with a residual charge.
Glassy carbon is known to be a highly polarizable material and the
electrostatic charge of lysine under these conditions lead to
strong retention of this compound. It is also interesting to note
that the order of migration (least to most retained), Lys-Phe-Thr,
changes to Lys-Thr-Phe between 600 and 800.degree. C. This change
in selectivity can be attributed to the increased
sp.sup.2-character and polarizability of the stationary phase
processed to 800.degree. C. in comparison to the 600.degree. C.
plate. Threonine, which contains a hydroxyl substituent, becomes
more strongly retained at 800.degree. C. These results illustrate
the ability of the glassy carbon nanofibrous stationary phase to
have tunable selectivity depending upon processing temperature.
[0150] Table 5 presents the plate numbers obtained using the glassy
carbon UTLC devices in comparison to previously published data.
TABLE-US-00009 TABLE 5 Plate numbers obtained for analysis of amino
acids using each electrospun glassy carbon UTLC plate and
comparison to a mixed stannic arsenate/cellulose phase. Plate
Number, N Compound 600.degree. C. 800.degree. C. 1000.degree. C.
Cellulose.sup.[7] Lysine 37500 .+-. 4500 6800 .+-. 650 330 .+-. 40
370 Threonine 195000 .+-. 6100 32400 .+-. 3400 330 .+-. 20 2100
Phenylaline 476000 .+-. 7900 29600 .+-. 4500 290 .+-. 30 N.A.
The 600 and 800.degree. C. carbon devices show a vast increase in
efficiency when compared to results using the mixed cellulose phase
described in. The 600.degree. C. devices result in plate numbers in
excess of 100,000 for threonine and phenylalanine over a
development distance of 2.0 cm. As was the case with the test
analytes, minimal band broadening occurred and typical developed
spot widths were 2.5 mm or under for the 600.degree. C. and
800.degree. C. electrospun carbon UTLC devices. It is also worthy
to note that highly efficient analyses were obtained using the
glassy carbon UTLC technique using development times of 5 minutes
and development distances of 3.5 cm or less while the comparison
work used a development distance of 15 cm. The electrospun glassy
carbon UTLC devices described in this study take a step towards one
of the overall goals of UTLC which is to improve performance while
reducing analysis time.
[0151] The 1000.degree. C. plate showed extreme band broadening and
was clearly not optimal for the essential amino acid analysis. In
order to examine the cause of the poor performance of the
1000.degree. C. plates, the sample capacity of each of the
electrospun glassy carbon UTLC devices was investigated. It was
found that the 600 and 800.degree. C. plates performed
satisfactorily in the concentration range of 10.sup.-7 to 10.sup.-5
M while the 1000.degree. C. device was overloaded at concentrations
higher than 10.sup.-6 M.
[0152] Previous work by both Poole and Guiochon demonstrated that
in capillary flow, small particle (<10 .mu.m) TLC and HPTLC, the
plate height, H, increases as a function of development distance.
For HPTLC using a support of 5 .mu.m diameter silica particles,
plate height was found to increase from .about.50 to 130 .mu.m
between development distances of 4 and 10 cm. For all analytes
studied using electrospun glassy carbon UTLC devices, the plate
height either decreases or remains constant over the range of
development distances (1-6 cm). The most strongly retained (lowest
Rf value) amino acid, lysine, was found to have the most
significant decrease in plate height (from 20 .mu.m 418 to 10
.mu.m) over the development range studied. Threonine and
phenylalanine, which have similar Rf values, both showed similar
behavior as plate height decreased from 3 to 1 .mu.m over a
development distance range of 4-6 cm. To the best of our knowledge,
the only other report describing a decrease in plate height with
increased migration distance uses pressurized planar
electrochromatography (PPEC). This study reports a decrease in H
from .about.600 .mu.m to 100 .mu.m at migration distances of 5 and
70 mm, respectively. More detailed studies are currently being
performed to elucidate the mechanism of the enhanced performance
seen in the electrospun glassy carbon UTLC devices.
[0153] FIG. 34 demonstrates the resolution of separations performed
with the electrospun glassy carbon UTLC devices and a comparison to
previously published literature values. As was the case with the
separation of laser dye mixtures, increasing the processing
temperature (from 600-800.degree. C. in this case) increases
selectivity which results in higher resolution analyses. 600 and
800.degree. C. devices showed the most desirable chromatographic
performance for high efficiency separations of the ternary
essential amino acid mixture. The electrospun carbon UTLC provides
much higher N and resolution when compared to previous analyses.
The results suggest that the mixture could be completely resolved
at even shorter analysis times and development distances.
[0154] Electrospun glassy carbon nanofibrous stationary phases for
UTLC were developed. This technique showed very high plate numbers
(10,000s to 100,000s) and differing selectivity for the analysis of
a set of 6 laser dyes as well as complete resolution of ternary
mixtures of laser dyes and essential amino acids. Developed spot
widths were similar to initial sample spot sizes, demonstrating
that band broadening is very small using this technique. The
selectivity of the devices can be tuned by pyrolyzing the
electrospun nanofibers at different temperatures. The experimental
results are in agreement with data obtained using Raman
spectroscopy which shows an increase in the sp.sup.2-character of
the carbon material with increasing pyrolysis temperature leading
to more conjugation, stronger .pi.-.pi. interactions and higher
polarizability. The most strongly retained compounds were highly
aromatic or charged compounds which were able to interact strongly
with the glassy carbon stationary phase via multiple interaction
modes. The described technique allows for the fabrication and use
of a nanofibrous carbon stationary phase for UTLC which is
chemically and mechanically robust and was found to have
outstanding chromatographic performance.
[0155] While the invention has been described in connection with
what is presently considered to be the most practical and preferred
embodiments, it is to be understood that the invention is not to be
limited to the disclosed embodiment(s), but on the contrary, is
intended to cover various modifications and equivalent arrangements
included within the spirit and scope of the appended claims, which
are incorporated herein by reference.
* * * * *
References