Solid phase coatings for microextraction

Abstract

An extractive system, such as SPME, has an adsorptive phase in the form of a porous coating that has essentially vertical, mutually supporting, columnar structures with nanospaces at the boundaries of the grains.

Claims

1. An apparatus for adsorbing then desorbing analytes, the apparatus comprising: a needle substrate; a coating attached to the needle substrate, the coating being adsorptive and porous; the coating including columnar nanostructures; nanospaces between adjacent columnar structures; and a thickness of the coating is circumferentially constant and tapered along a length of the needle substrate, with longer columnar nanostructures at a point of the needle substrate, and becoming thinner with shorter columnar nanostructures as a distance from the point increases.

2. The apparatus of claim 1, wherein the coating comprises zirconia and a phosphonate, or zirconia and a perflourinated phosphonate, or both.

3. The apparatus of claim 1, wherein the coating comprises a metal and a phosphonate, or a metal oxide and a phosphonate, or both.

4. The apparatus of claim 1, wherein the columnar nanostructures are mutually supporting.

5. The apparatus of claim 1, wherein the columnar nanostructures are essentially vertical relative to the needle substrate at a location of attachment.

6. The apparatus of claim 1, wherein the coating includes silicon, or carbon, or both.

7. The apparatus of claim 1, wherein the coating includes silica.

8. The apparatus of claim 1, wherein the coating comprises a silanized silica, or a silanized silicon, or both.

9. The apparatus of claim 1, wherein the coating includes a polymeric material.

10. An apparatus for adsorbing then desorbing analytes, the apparatus comprising: a substrate; a coating attached to the substrate; the coating being adsorptive and porous; the coating including mutually supporting, columnar nanostructures and nanospaces between adjacent columnar structures; and a thickness of the coating being circumferentially constant.

11. The apparatus of claim 10, wherein the coating comprises zirconia and a phosphonate, or zirconia and a perflourinated phosphonate, or both.

12. The apparatus of claim 10, wherein the coating comprises a metal and a phosphonate, or a metal oxide and a phosphonate, or both.

13. The apparatus of claim 10, wherein the columnar nanostructures are essentially vertical relative to the substrate at a location of attachment.

14. The apparatus of claim 10, wherein a surface of the coating is hydrophobic.

15. The apparatus of claim 10, wherein a tape adhesive test, in which a piece of tape is pasted on the coating then removed, results in a SEM showing no damage to the columnar nanostructures.

16. An apparatus for adsorbing then desorbing analytes, the apparatus comprising: a substrate; a coating of columnar nanostructures attached to the substrate, the coating being adsorptive and porous; boundaries between adjacent columnar nanostructures are spaced but close enough to be mutually supporting; and a thickness of the coating is tapered along a length of the substrate.

17. The apparatus of claim 16, wherein the columnar nanostructures are essentially vertical relative to the substrate at a location of attachment.

18. The apparatus of claim 16, wherein a thickness of the coating is circumferentially constant.

19. The apparatus of claim 16, wherein the coating comprises aluminum, carbon, silicon, silica, titanium, zirconium, or combinations thereof.

20. The apparatus of claim 16, wherein the coating comprises aluminum, carbon, silicon, titanium, zirconium, or combinations thereof.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) FIG. 1. SEM images of silica fibers sputtered with silicon at 4 mTorr. The fibers were positioned in the direction of the sputter target, but 1.5 cm off of the center of the target. Sputtering times were 60 min: (a) side view, (b) top view, (c) profile view, 120 min: (d) side view, (e) top view, and (f) profile view, and 180 min: (g) top view, (h) profile view. NOTE: in image (c), the profile view represents the bottom part of the fiber. The bottom point refers to the furthest part from the target and hence has lowest thickness, whereas (f) and (i) represent the top part, nearest to the target.

(2) FIG. 2. Ratio of peak areas of straight chain alkanes obtained with commercial 7 μm PDMS fiber and ca. 1 μm present sputtered fiber. (Here, ‘7 μm’ and ‘1 μm’ refer to the thickness of the coatings on the fiber.)

(3) FIG. 3. Gas chromatograms of saturated alkanes (C.sub.8-C.sub.16) obtained with commercial 7 μm PDMS fiber and present ca. 1 μm sputtered fiber

(4) FIG. 4. Ratio of peak areas of primary alcohols obtained with commercial 7 μm PDMS fiber and present sputtered fiber ca. 1 μm (S-fiber)

(5) As with the alkanes, the separation of alcohols was performed at University of Tasmania where we got the same results (FIG. 5). Sputtered fibers were performing better than the commercial PDMS fibers, especially in the high mass regime. The fact that the present fibers were working in different labs and giving superior results than the commercial counterparts is encouraging and vouches for the robustness and usefulness of the product.

(6) FIG. 5. Primary alcohols (hexanol, heptanol, octanol, decanol and dodecanol) extracted using present sputtered ca. 1 μm three hour sputtered fiber and PDMS 7 μm fiber. Present sputtered fiber shows that it gives more response for the test mix used.

(7) FIG. 6. Head space extraction of components of a commercial beer sample using present sputtered and PDMS 7 μm fibers.

(8) FIG. 7. Head space extraction of components of a sea water extract using present sputtered ca. 1 μm and PDMS 7 μm fibers.

(9) FIG. 8. Extraction profile of present sputtered 1 μm fiber using , C.sub.8, C.sub.9, C.sub.10 and C.sub.12 primary alcohol using different extraction times : 1, 3, 5, 7, 10, and 15 minutes.

(10) FIG. 9. Ratio of peak areas of primary alcohols obtained with commercial 85 μm CAR-PDMS fiber, 65 μm DVB-PDMS and 1 μm present sputtered fiber.

(11) FIG. 10. Ratio of peak areas of saturated alkanes obtained with commercial 85 μm CAR-PDMS fiber, 65 μm DVB-PDMS and 1 μm present sputtered fiber.

(12) FIG. 11. SEM micrographs of sputtered silicon layer (1 hour sputtering) on silicon wafer before and after scotch-tape adhesion testing.

(13) FIG. 12. SEM images of ˜2.2 μm thick silicon coated present fibers.

DETAILED DESCRIPTION

Example

(14) Experimental

(15) The preparation of nanoporous silica coatings for SPME via sputtering consists of the following steps:

(16) I. Silica Fiber Preparation

(17) Polyimide-coated silica fibers (140 μm in diameter) were cut in lengths of 3.3 cm and the polyimide coating on the fibers was burned off. The resulting diameter of the fibers was 120-124 μm. The fibers were visually inspected for any left-over polyimide coating.

(18) II. Sputter Deposition

(19) Cut fibers were fixed onto the platen (sample holder) of a PVD 75 (sputter system from the Kurt J.Lesker Co.) in a way such that the fibers hang nearly vertically in the chamber. Silicon (99.999%) was DC magnetron sputtered at 4 mTorr and 200 W power. Argon was employed as the sputtering gas. Sputtering was done for different times to study the effect of thickness on the extraction capacities of the fibers.

(20) III. Hydroxylation of the Silica Surface

(21) Sputtered fibers were treated in piranha solution (7:3::H.sub.2SO.sub.4:H.sub.2O.sub.2) at 85-90° C. for 45 minutes to introduce additional silanol groups on the sputtered silicon surface. This was followed by rinsing these fibers with ultra-high purity water and drying them in nitrogen for 15 minutes

(22) IV. Rendering the SPME Silicon Coating Hydrophobic

(23) Silanization chemistry was used to introduce C.sub.18 chains on the silica surface via chemical vapor deposition in a home-made oven. A liquid phase deposition of the silane should also be possible. The fibers were placed in the oven chamber, which was evacuated to a base pressure of 0.5 Torr. After that, 0.1 mL of n-octadecyldimethylmonomethoxysilane(C-18 silane) was directly injected into the preheated oven at 200° C. The vapors of the silane were allowed to remain in the chamber for 20 min. Finally, the chamber was purged three times to remove the unreacted C.sub.18 silane.

(24) V. Attachment of SPME Fibers

(25) Finally, using epoxy glue (EPO-TEK 353ND-T), the fibers were attached to the plunger needle of SUPELCO SPME assemblies. The final length of the exposed fibers was 1.0 cm.

(26) While sputtering silicon onto the fibers, witness silicon wafers were also sputtered so they could be characterized using XPS and water contact angle goniometry.

(27) VI. GC-FID Conditions

(28) The test mix consisted of 1 ppm each of decane (C.sub.10), dodecane (C.sub.12). tetradecane (C.sub.14), and hexadecane (C.sub.16) in water. The solution was made by dissolving 2 μL each of C.sub.10, C.sub.12, C.sub.14, and C.sub.16 in 10 mL of ethanol and then diluting 50 μL of this stock solution with 10 mL of ultra-high purity water to attain 1 ppm concentrations. For analysis, 5 mL of this solution was taken in a 20 mL SPME vial. The stock solution was stored at 4° C.

(29) The GC column used was HP-5 5% phenyl methyl siloxane with capillary dimensions of 30.0 m×250 μm×0.25 μm. Fibers were preconditioned at 280° C. for 360 min. However, sometimes longer times were needed to precondition the fibers. GC conditions were: incubation time of the analyte solution in the vial: 5 min. at 40°C., headspace extraction time of the analytes: 10 minutes at 40° C., desorption conditions in the GC injection port: 280° C. for 1 minute. The initial column temperature was 70° C. with a ramp rate of 20° C./min up to 200° C. followed by a ramp rate of 30° C./min to 300° C., with a hold for 3 min at 300° C. The total run time for the analysis was 13.83 min. The fibers were baked for 10 min between the runs at 280° C.

(30) Results

(31) Silica coatings for SPME were prepared by sputtering silicon. This example was performed on silica fibers, but various other metal fibers could be used to enhance the mechanical stability of the fiber. The substrates were hanged vertically in the chamber, approximately perpendicular to the target. The fibers, after sputtering of silicon at 200 W DC magnetron power, were observed via SEM (FIG. 1). The resulting coatings were quite porous with vertical columnar structures.

(32) This could be explained because of the lower mean free path of the sputtered silicon atoms due to a relatively high pressure in the chamber (4 mTorr) and relatively long throw distance (17-20 cm) (distance between the target and the fiber substrates). A lower mean free path means that the silicon atoms undergo multiple collisions before reaching the substrate and hence lack directionality. We calculate the mean free path of silicon atoms at 4 mTorr and other pressures using the following well-known equation:

(33) $\lambda =\frac{k_{B}T}{\sqrt{2}{\pi d}^{2}p}$
where, λ=mean free path, k.sub.B=Boltzmann constant, T=temperature in K (298 K), d=diameter of the molecule (diameter of silicon atom is 222 pm.), p=pressure in Pascal. Obviously, the application of this equation isn't quite correct here. We have assumed that all the atoms in the chamber are silicon, which is not the case. Nevertheless, this equation should provide a fairly accurate measure of the mean free paths of the present sputtered silicon atoms.

(34) TABLE-US-00001 TABLE 1 Mean free paths of Si atoms at various pressures. Sputtering pressure (mTorr) Mean free path of Si atom (cm) 4 3.53 3 4.70 2 7.05 1 14.10

(35) Table 1 shows the mean free path of Si atom at various sputtering pressures in the PVD 75 chamber. It is evident that the mean free path at 4 mTorr is rather small compared to the throw distance, hence leading to the vertical columnar structures observed on the present fibers instead of the type of structures that would be obtained in an oblique angle deposition.

(36) Under the above deposition conditions, the vertical columnar structures showed some tapering as the thickness of the fiber decreased from top to bottom. For a 1 cm fiber sputtered for 3 h, the thickness was ca. 1.1 μm at the bottom of the fiber (closest to the target) and 0.86 μm at the top (furthest from the target). The tapered thicknesses from top to bottom were also seen with 2 h and 1 h sputtered fibers. Overall, the vertical columnar structures seem to be beneficial as they are robust and showed good extraction capabilities.

(37) After sputtering, the silicon fibers were treated with piranha solution. The rationale for treatment with this solution was to introduce additional silanol groups onto their surfaces, which in turn, would provide more binding sites for C.sub.18 silane, and hence greater hydrophobicity. The C-18 silane was vapor deposited in a home made oven. Spectroscopic ellipsometry showed an increase in thickness of about 1.2 nm on a witness (non-sputtered) silicon wafer confirming the attachment of the C.sub.18 silane.

(38) As the thicknesses of the sputtered coatings increased, so did their extraction capacities. Therefore, comparatively thicker coatings were made by sputtering silicon for a longer period of time, e.g., three hours (as opposed to 1 or 2 hours) at 4 mTorr. Two of these fibers (fibers I & II) were tested using GC-FID (gas chromatography-flame ionization detector) analysis/detection and were compared to 7 μm PDMS commercial fibers. The increase in thickness of the sputtered silicon fibers enhanced extraction capacities, i.e., more signal was obtained from the thicker coatings. For higher molecular weight compounds, C.sub.14 and C.sub.16, the extraction efficiencies were better with the present fiber as compared to commercial PDMS 7 μm fibers (FIG. 2).

(39) A mixture of alkanes (C.sub.8-C.sub.16) was separated at the University of Tasmania with the same trends. The sputtered fibers (˜1 μm) outperformed the PDMS (7 μm) fibers, especially in the high mass regime (C.sub.13-C.sub.16) (FIG. 3).

(40) The sputtered fibers were also used for separating a mix of primary alcohols, where they showed greater selectivity for the higher molecular weight analytes than the commercial PDMS 7 μm fiber (see FIG. 4). The extraction conditions were the same as used for extracting the alkanes. The results suggested that the present fibers had unique selectivity for polar alcohols compared to PDMS fibers.

(41) In addition, these fibers were also used for some real world samples, including sea water extract and beer analysis. ‘James Boag's Premium Light’ Launceston, Australian beer was analyzed. The extraction conditions were: sample volume: 5 mL beer directly from the bottle; extraction temperature: 45° C.; extraction time: 10 minutes; desorption temperature: 280° C.; desorption time: 1 minute; initial column temperature: 60° C., ramp 10° C./min till 240° C., ramp 20° C./min till 280° C. and hold for 3 minutes.

(42) No attempts were made to identify the peaks; rather qualitative comparisons were made between the chromatograms obtained from the present sputtered fiber and the 7 μm PDMS counterpart. FIG. 6 shows that the present fiber outperformed the commercial PDMS 7 μm fiber, especially in the high mass range region.

(43) We also analyzed a sea water extract. Tens of gallons of sea water were passed through a C.sub.18 solid phase extraction (SPE) cartridge and the eluent was dried. The dried extract was dissolved in 80:20 water: methanol mixture. 25 μL of this solution was added to 5 mL of ultra-pure water and used for analysis. The extraction conditions were: sample volume: 5 mL; extraction temperature: 40° C.; extraction time: 8 hours; desorption temperature: 280° C.; desorption time: 1 minute; initial column temperature: 60° C., ramp 3° C./min till 240° C., ramp 20° C./min till 280° C. and hold for 3 minutes.

(44) FIG. 7 compares the results of analysis of sea water extract using sputtered fiber (˜1 μm) and PDMS fiber (7 μm) which depicts some interesting findings. BYU fiber was able to extract some peaks that the commercial PDMS fiber was not able to extract (see oval). It is important to mention that these peaks were in the low mass regime, indicating unique selectivities of the present fiber, presumably due to unreacted —OH groups and C18 chains decorating the surface. Blue (dotted) ovals show higher extraction and unique peaks extracted by the present fiber in the mid-mass range regime. The rise in baseline around 450 second in case of the sputtered fiber indicates that higher quantities of analytes were extracted as compared to PDMS fiber. 2-D GC (GC×GC) would be tried in the future to isolate all the separate compounds.

(45) To understand why the present fibers perform better in the high mass regime, it is important to understand the mechanism of action of the present fibers. While not limited to a theory, PDMS is an example of liquid stationary phase, which extracts analytes via absorption. The amount of analyte extracted depends on the concentration of analyte in the sample matrix and follows a linear trend. On the other hand, solid coatings (for example the present sputtered fibers) appears to work on the principle of adsorption, that is analytes stick on specific sites on the coatings and as there are finite number of sites, the extraction process becomes competitive at higher concentrations.

(46) While not limited to a theory, assume that the solid coating has a finite number of ‘sites’ on the surface. A low molecular weight analyte—‘analyte.sub.1’ and high molecular weight analyte—‘analyte.sub.2’ are being extracted:
Analyte.sub.1+Site.Math.Analyte.sub.1−site (equilibrium constant=K.sub.1   (1)
Analyte.sub.2+Site.Math.Analyte.sub.2−site (equilibrium constant=K.sub.2)   (2)
(analyte.sub.1-side and analyte.sub.2-site refer to complexes between the respective analyte and site)

(47) As analyte.sub.2 is higher molecular weight, its extraction would be enthalpy favored and, therefore, under the extraction conditions used K.sub.2>K.sub.1 Reversing (1) and adding to (2):
Analyte.sub.1−site+Analyte.sub.2.Math.Analyte.sub.2−site+Analyte.sub.1   (3)

(48) The above equation describes how a higher molecular weight specle would displace lower molecular weight analytes. The equilibrium constant of the displacement reaction in (3) is given by K.sub.2/K.sub.1.

(49) To demonstrate that the present sputtered fibers follow an adsorption mechanism, we created an extraction profile separating a mixture of octanol, nonanol, decanol and dodecanol (1 ppm each in water). The mixture was extracted at 40° C. for 1, 3, 5, 7, 10, and 15 minutes. The peak areas for different analytes were recorded as a function of extraction time (see FIG. 8). Extraction conditions were: sample volume: 5 mL; extraction temperature: 40° C.; extraction time: variable; desorption temperature: 280° C.; desorption time: 1 minute; initial column temperature: 70° C., ramp 10° C./min till 200° C., ramp 30° C./min till 300° C. and hold for 3 minutes.

(50) It is evident from FIG. 8 that the initial response for dodecanol was quite small at the 1 minute extraction time. As the duration of extraction increased, the response of dodecanol increased significantly at the expense of loss of response of other lower molecular weight analytes. Therefore, the sputtered fibers can be used for two types of extraction:

(51) (1) Short time extraction: extract lower molecular weight species

(52) (2) Longer time extraction: extract higher molecular weight species

(53) It is worth emphasizing that even though PDMS has a different mechanism of extraction than the present sputtered fibers, it is one of the most commonly used commercial extraction phase. Hence, a comparison of PDMS with the present fibers is reasonable. Nevertheless, we have extended this comparison to other solid extraction phases— (1) Carboxen-PDMS (CAR-PDMS) 85 μm fiber a. Used for low molecular weight analytes (ii) PDMS-DVB 65 μm fiber a. Used for high molecular weight analytes

(54) We compared the response of the present sputtered fiber with the above mentioned solid adsorption coatings using a mixture of alcohols and alkanes.

(55) The mixture of alcohols consisted of C.sub.7, C.sub.8, C.sub.9, C.sub.10 and C.sub.12 primary alcohols (1 ppm in water). The separation conditions were similar as employed earlier for separation of alcohols, except that the desorption temperature was kept at 260° C. for all the fibers and the extraction time was 3 minutes. The desorption temperature value was influenced by the upper operating temperature limit for comparison adsorbent fibers. FIG. 9 compares the response of ˜1 μm thick sputtered fiber with 85 μm CAR-PDMS and 65 μmDVB-PDMS fibers.

(56) For C.sub.7 alcohol the PDMS-DVB gave 58 times and CAR-PDMS gave 45 times the response of sputtered fiber. Considering that extraction profiles scale down iinearly with thickness, we are actually performing better than the other two fibers. The ratios become more favorable as we go to higher molecular weight analytes. Even for C.sub.8 alcohol, CAR-PDMS only does 3 times better than sputtered fibers, despite being 85 times thicker. For C.sub.10, we outperform the 85 times thicker CAR-PDMS and give ⅓ response as the 65 times thicker DVB-PDMS. The most interesting finding was that CAR-PDMS was not able to extract C.sub.12 alcohol at all. Overall, for higher molecular weight compounds, we had comparable and in some cases better response than the coatings that were 85 times thicker than ours.

(57) After outperforming thicker adsorbent coatings in separating alcohols, we tested the alkanes. We have a unique selectivity for alcohols, other fiber might not. The comparison fibers in question are used extensively for extraction of hydrophobic compounds, like aikanes. Therefore, a comparison for separating alkanes would be valuable. The separation conditions were same as mentioned above in case of alcohols.

(58) FIG. 10 shows the comparison of the present fiber with two other commercial adsorbent coatings for the extraction of a mixture of C.sub.10, C.sub.12, C.sub.14, and C.sub.16 alkanes. The results follow the same trend. For C.sub.10, CAR-PDMS gave 212 times and PDMS-DVB gave 161 times the response of the present fiber, which was expected as the present fibers have shown to work better for higher molecular weight analytes. The ratios decrease for C.sub.12 and for C.sub.14, CAR-PDMS gave 1.2 times and DVB-PDMS gave 3.2 times the response of the present fiber. Further, for C.sub.16, the present 1 μm thick fiber outperformed the 85 μm thick CAR-PDMS fiber with ease, giving 5 times the response. 65 μm thick DVB-PDMS gave 1.3 times the response of the present fiber.

(59) Also, the present sputtered fibers, were free of any carryover effects, whereas, the other two solid adsorbent coatings had significant carryover effects.

(60) To show the robustness of the present sputtered films, we performed a scotch-tape adhesive test. This is one of the yard-sticks to measure the adhesion of the films to the substrates. We sputtered Si using the same instrument, conditions and geometry onto planar silicon wafers for 1 hour and did SEM on it (see FIG. 11 (Before)). A piece of scotch tape was pasted on the surface and then removed. The residue leftover from the tape was cleaned by sonicating the wafer in a vial full of acetone. The SEM after the test showed no damage to the structures (see FIG. 11 (After)).

(61) As the thickness of the sputtered coatings increases, so did the response. On the same lines, in order to enhance the extraction capabilities, we have sputtered a ˜2.2 μm silicon coating on silica fibers. FIG. 12 shows the SEM of the present thicker coatings. As expected, these thicker coatings showed higher extraction capacities (larger signals) compared to their thinner counterparts.

Conclusion

(62) The SPME coating prepared by sputtering can provide robust coatings with high porosities. The method could be applied to produce various metal or metal oxides coatings. Moreover, the selectivity of the coating could be manipulated using different chemistries on the surface (example silanization chemistry on silica coatings to yield desired selectivities). Sputtering provides a better control over coating thickness, with high reproducibility. The present sputtered fibers were able to outperform thicker liquid and solid coatings, especially in the high mass regime, providing huge potential for a faster analysis.

(63) While this invention has been described with reference to certain specific embodiments and examples, it will be recognized by those skilled in the art that many variations are possible without departing from the scope and spirit of this invention, and that the invention, as described by the claims, is intended to cover all changes and modifications of the invention which do not depart from the spirit of the invention.