2195
Anal. Chem. 1984, 56, 2195-2199
of 160000 (27) is close to the effective molecular weight cutoff of 200 000 for this membrane, there is a clogging effect of the pores with longer time periods. This could explain the decrease in the CL signal observed for the membrane with smaller pores, while the signal for the 0.2-pm pore size membrane was constant. In addition, the CL signals observed were over 10 times larger when using the 0.2-pm pore nylon membrane as opposed to the 0.04-pm pore poly(propy1ene) membrane. Because a positive displacement syringe pump was used to control the flow rate through the membrane, that total flow rate would not have been different for the two pore sizes. However, the flow profiles through the pores and at the membrane surface are probably quite different for the two membranes. This change in flow characteristics is no doubt responsible for the large change in signal intensity. Due to the larger signals and the lack of any time-dependence problems for longer analysis, the 0.2-pm nylon pore size membrane was chosen for future glucose analyses.
CONCLUSION Although use of Cu(phen)32+yields lower detection limits, HRP has better potential since it requires milder pH and does not suffer the complication of a time-dependent response. Continuous-flow FIA is more advantageous than stopped-flow FIA on the basis of detection limits, linear range, convenience, and throughput (100 samples/h vs. about 15 samples/h). Although both membranes yield good precision for experiments of under an hour duration, the 0.2-pm pore membrane yields superior precision and signal intensity for experiments of longer duration; this behavior is probably related to the enzyme size and may not be identical for other enzymes. Registry No. HRP,9003-99-0; Cu(phen)32f,20243-47-4; EC 1.1.3.4,9001-37-0; glucose, 50-99-7; luminol, 521-31-3;hydrogen peroxide, 7722-84-1.
LITERATURE CITED Kelly, T. A.; Christian, G. D. Anal. Chem. 1981, 53, 2110-2114. Guilbault, G. C.; Lubrano, 0. T. Anal. Chlm. Acta 1973, 64, 439-455. Updlke, S. J.; Hicks, G. P. Nature (London)1967, 214, 986-988. Thevenot, D. R.; Coulet, P. R.; Sternberg, R.; Laurnet, J.; Gautheron, C. Anal. Chem. 1979, 51, 96-100. Seitz, W. R. Methods in Enzymol. 1978, 57, 445-462. Auses, J. P.; Cook, S. L.; Maloy, J. T. Anal. Chem. 1975, 47, 244-249.
Bostick, D. T.; Hercules, D. M.. Anal. Lett. 1974, 7 , 347-353. Bostlck, D. T.; Hercules, D. M. Anal. Chem. 1975, 47, 447-452. Williams, D. C.; Huff, G. F.; Seitz, W. R. Clin. Chem. (Winston-Salem, N . C.) 1978, 22, 372-374. Rldder, C.; Hansen, E. H.; Ruzicka, J. Anal. Lett. 1982, 15, 1751-1766.
Pllosof, D.; Nleman, T. A. Anal. Chem. 1982, 54, 1698-1701. Nau, V. J.; Nieman. T. A. Anal. Chem. 1979, 51, 424-428. Pilosof. D.; Nleman, T. A. Anal. Chem. 1980, 52, 662-665. Pilosof, D.; Malavolti, N. L.; Nleman, T. A. Anal. Chim. Acta, submitted for publication. Puget, K.; Mlchelson, A. M. Blochlmle 1976, 58, 757-7513, Mlsra, H. P.; Squatrlto, P. M. Arch. Biochem. Biophys. 1982, 215, 59-65.
Freeman, T. M.; Seitz, W. R. Anal. Chem. 1978, 50, 1242-1246. Betteridge, D. Anal. Chem. 1978, 50, 832A-846A. Ranger, C. 0. Anal. Chem. 1981, 53, 21A-30A. Klopf, L.; Nleman, T. A. Anal. Chem. 1983, 55, 1080-1083. Rule, G., Seitz, W. R. Clln. Chem. (Wlnston-Salem, N . C . ) 1979, 25 (9), 1835-1638. Streitwiser, A., Jr.; Heathcock, C. H. ”Introduction to Organic Chemistry”; Macmillan: New York, 1976; Chapter 25. Laitlnen, H. A.; Harris, W. E. “Chemical Analysis”, 2nd ed.; McGrawHill: New York, 1975; pp 404-405. VanDyke, D., unpublished results, University of Illinois, 1982. Cormier, M. J.; Prichard, P. M. J . Bid. Chem. 1968, 243, 4706-4714.
Chen, T.; Nekimken, H., unpublished results, University of Illinois, 1983.
Tsuge, H.; Natsuakl, 0.; Ohashi, K. J . Biochem. (Tokyo) 1975, 78, 835-843.
RECEIVEDfor review April 20,1984. Accepted June 4,1984. This research was supported by the National Science Foundation (Grant CHE-81-08816).
Comparison of Phase-Resolved and Steady-State Fluorimetric Multicomponent Determinations Using Wavelength Selection Linda B. McGown* and Frank V. Bright
Department of Chemistry, Oklahoma State University, Stillwater, Oklahoma 74078
Results for the fluorimetric determination of a three-component system (anthracene, POPOP, and Me,POPOP) using phase-resolved fluorescence spectroscopy to combine fluorescence iifetlme selectivity with wavelength selectivity are compared with results obtalned for conventional steadystate determinations uslng wavelength selectivity alone. A 4-fold Increase In accuracy and a IO-fold decrease In error magnitude are achieved for the phase-resolved approach relative to the steady-state results. The comblnatlon of Iifetime with wavelength selectivity requires no addltlonal measurements or calculations when phase-resolved fluorescence Is used to Implement the llfetime selectivity, once the optlmal condltlons for the method have been determined.
lifetime with wavelength selectivity is demonstrated in the work described here for a three-component system. The theory and instrumentation of PRFS have been described elsewhere (1,2)and are based on the phase-modulation technique for determination of fluorescence lifetimes ( 3 , 4 ) . Briefly, the sample is excited with sinusoidally modulated light having an angular frequency w and intensity E ( t ) a t time t
E ( t ) = E,(1
+ me, sin ut)
(1)
where Eorepresents the dc intensity and me, is the degree of modulation (the ratio of the peak ac intensity to the dc intensity). The resulting fluorescence F(t) of a single component sample with exponential decay will be phase-shifted by angle and appear as F ( t ) = F,(1 me, cos 4 sin ( u t - $)) (2)
+
The use of phase-resolved fluorescence spectroscopy (PRFS) to improve the accuracy of multicomponent fluorescence determinations by combining selectivity based on fluorescence
where F, represents the dc component of the fluorescence intensity. For a solution containing more than one fluorescent species, the observed fluorescence will be the sum of the individual contributions given by eq 2, each species having
0003-2700/84/0356-2195$01.50/00 1984 American Chemical Society
2196
ANALYTICAL CHEMISTRY, VOL. 56, NO. 12, OCTOBER 1984
a characteristic phase shift $ and dc contribution Fo. In PRFS, the fluorescence resulting from excitation of the form shown in eq 1 is observed with phase-variable detection. For a solution containing two components A and B, the phase-sensitive signal a t detector phase $D will be where k A and kg are constants that depend upon the spectral characteristics of the species. Applications of PRFS described in the literature have primarily involved the selective nulling of each component in two-component mixtures (2,5). A component is nulled by setting the detector phase 90° from the component phase angle (e.g., $D = $A f 90’ to null component A), so that the phase-resolved intensity of the component is zero and the remaining, nonnulled component can be measured. Recently, improved accuracy for determinations of two fluorescent components with essentially identical emission and excitation spectra using measurements at two nonnulling detector phase angles relative to the nulling approach was demonstrated (6). For discussion purposes, the detector phase angle that is in phase with a component and therefore yields the maximum magnitude of phase-resolved intensity for the component shall be referred to as the “phase angle” of the component. The detector phase angle exactly 90° out of phase with the component, a t which the phase-resolved intensity is zero, is referred to as the “null phase angle” of the component. In the work described here, results for the analysis of a three-component system using conventional steady-state measurements at sets of three different emission wavelengths are compared with results for the same system using phaseresolved fluorescence measurements at sets of three different combinations of wavelength with detector phase angle. Anand thracene, POPOP (1,4-bis(5-phenyloxazol-2-yl)benzene), benzene) Me,POPOP (1,4-bis(4-methyl-5-phenyloxazol-2-yl) were chosen for the model three-component system because of their spectral overlap and relative lifetime differences. The results show that, although the three-component system can be successfully analyzed by using steady-state wavelength selectivity alone, the accuracy of the determinations is improved as much as 4-fold when lifetime selectivity is used in conjunction with wavelength selectivity, and the magnitudes of the errors can be reduced by an order of magnitude. Results are also shown for determinations of the threecomponent system by using PRFS measurements at constant wavelength a t (1)the null phase angles of each of the components and (2) a t three optimal nonnull detector phase angles. These results demonstrate the superior accuracy of the nonnull detector phase angle approach for the three-component system.
EXPERIMENTAL SECTION Standard Of popop MezPOPOP (Aldrich),and anthracene (Eastman) were prepared by adding the appropriate weight of the component to 100.0 mL of absolute ethanol (U.S. Industrial Chemicals Co.) and sonicating for 30 min, followed by a 100-folddilution with ethanol. Mixtures were prepared by combining the appropriate volumes of the standard solutions of the components with no further dilution, All fluorescence measurements were made by using disposable polyethylene cuvettes (Precision Cells, Inc.). Data Collection. All fluorescence measurements (steady-state and phase-resolved) were made with a 450-W xenon arc lamp source and photomultiplier tube (Hamamatsu R928) detection. A frequency Of 30 MHz was for PRFS measurements, which were all taken in delta phase mode (an instrumental mode which provides simultaneous comparison of the sample beam &h a podion of the modulated excitation beam that has been diverted to a reference cuvette and detected by a phase-sensitive reference PMT). Steady-state measurements were
similarly made in a ratiometric mode. For PRFS determinations, all solutions were measured first at one detector phase angle and then at the next, and so on, to minimize the effect of imprecision in setting the phase angle. Blank contributions to measured intensities due to the ethanol solvent were negligible so that no correction of the measured intensities was necessary. Solutions were not degassed for these experiments. All measurements (steady state and PRFS) were made in triplicate, each measurement being the electronic average obtained by integration of 100 samplings over approximately 30 s, performed internally by the SLM instrument. Data Analysis. For all determinations, a square 3 x 3 matrix was generated by measurement of solutions under three different sets of conditions. For steady-statedeterminations, three different emission wavelengths were used. For PRFS determinations, three different combinations of emission wavelength with detector phase angles were used. In either case, three independent equations were generated: condition set 1: condition set 2: condition set 3: I3
= 7.4,3CA + f P P , 3 c P+ f M , 3 c M
(4)
in which the abbreviations A, P, and M are used to denote anthracene, POPOP, and Me2POPOP, respectively. The I values are the fluorescence intensitigs (steady state or PRFS) of the solutions being analyzed. The I values are the molar fluorescence intensities (steady state or PRFS) of each component, determined by measuring a standard solution of each component at each set of conditions. The I and f values were entered by hand into a Apple IIe microcomputer, and the generated matrix was solved for CA, C p , and C M using Cramer’s rule. In this work, four different solutions were analyzed. One solution contained all three components, and three solutions were binary mixtures. The binary mixtures were treated as threecomponent systems by using the 3 X 3 matrix format to test the ability to determine the absence of a possible component (i.e., zero concentration). Component concentrations were adjusted to give approximately equal intensity contributions from each of the components present.
RESULTS AND DISCUSSION Steady-State Fluorescence Spectra. The steady-state fluorescence excitation and emission spectra (uncorrected) of anthracene, POPOP, and MezPOPOP are shown in Figures 1 and 2, respectively. Overlap of the excitation spectra is extensive, and selectivity based on wavelength was achieved by varying the emission wavelength only. Four emission wavelengths were used, including the emission maximum of each compound (381 nm for anthracene, 417 nm for POPOP, and 429 nm for Me2POPOP) and 453 nm a t which the intensity contribution of anthracene is low relative to both the POPOP and MezPOPOP contributions. Phase-Resolved Fluorescence Lifetimes. Graphs of phase-resolved intensity as a function of detector phase angle are Shown for ~ ~ component c h in Figure 3. The longer lifetime of anthracene is evidenced by the occurrence of its maximal intensity at a larger (later) detector phase angle than the maxima for POPOP and MezPOPOP which have similar, shorter lifetimes. When popop is used as a reference for lifetime determinations (instead of a scattering solution) as has been recommended (7), POPOP is assumed to have a lifetime of 1.35 ns (7) and anthracene and MezPOPOP lifetimes are calculated to be 1.42 and 4.07 ns, respectively. These values are in good agreement with reported values of 1.45 ns for MezPOPOP in ethanol (7) and 4.26 ns for anthracene in benzene (8).
ANALYTICAL CHEMISTRY, VOL. 56,
NO. 12, OCTOBER
1984
2197
Table I. Precisons of Phase-Resolved Intensity Measurements av std deviation'
detector phase angle, deg 207.0 183.8 184.5
180.0 164.4 156.3 148.2 140.1 132.0
(anthracene null) (POPOP null) (Me,POPOP null)
abs
rel, %
9.1 12 6.7 12 9.7 16 24 17 14
1.0 18 4.3 2.1 0.31 0.37 0.47 0.32 0.25
'Values shown are averages for seven solutions (three standards and four mixtures) measured in triplicate at each detector phase angle.
332
300
364
428
396
EXCITATION WAVELENGTH ( n d Flgure 1.
Me,POPOP, PM.
Steady-state excitation spectra (A,, = 401 nm): (-) 1.22 pM; (- -) POPOP, 1.29 pM; (. .) anthracene, 1.O2
-
-
...........
I
365
397
429
461
525
493
WISSION WAVELENGTH (nm)
Figure 2. Steady-state emission spectra
(Ae, = 360 nm). Legend is
the same as in Figure 1 for solutions.
0
45
90
135
180
225
270
315
360
DETECTOR PHASE ANGLE (de8rm.e)
Flgure 3. Phase-resolved fluorescence intensity as a function of de,, = 355 nm, ,A, = 400 nm. (0)Me,POPOP, tector phase angle. A 1.22 pM; (V)POPOP, 1.29 pM; (0)anthracene, 1.02 pM.
Choice of Detector Phase Angles for PRFS Determinations. Considerations for choosing detector phase angles have been previously discussed (6) and involve a compromise between maximum discrimination between components (achieved at the null phase angles of the components) and maximum precision (achieved at the phase angles of the components, a t which maximum intensities are obtained). It has been shown that measurements midway between these two, i.e., 45O out of phase with the components, can achieve an effective compromise between accuracy and precision for determination of components with similar fluorescence spectral and lifetime characteristics (6). Determinations were performed at constant wavelength (Aex = 355 nm, A, = 400 nm) to compare the PRFS determinations for the three-comonent system using the null phase angles of each component with determinations using nonnull phase angles. Measurements were made for each of the four solutions (one ternary and three binary mixtures, as discussed above) and for the three standard solutions, at the three null phase angles and a t a total of six nonnull phase angles. Average precisions for measurements made a t each angle are shown in Table I. Both absolute standard deviations and relative values are given, from which it can be seen that the absolute precisions are relatively constant at all angles. Therefore, the relative precisions will be worst at the null phase angles a t which the total signal is smallest and will improve with increasing distance from the null angles. Eventually, a maximum precision is reached which covers a broad range of angles, including the phase angles of the components. The results of the null and the nonnull phase angle determinations are shown in Table 11. For the nonnull determinations, results are shown for the combination of three angles from the total six used that gave the best results. Since wavelength selectivity was not used in these determinations, neither set of results is very good. However, it is clear that the accuracy obtained by using nonnull phase angles is better than the accuracy obtained by using null phase angles. This is due to the relatively larger uncertainty of the null phase angle measurements as shown in Table I, as well as to the difficulties in locating the exact null phase angle of a component (6). Note that the results in Table I1 are summarized in two ways. First, the average relative errors are shown for each set of determination conditions. The averages of the absolute values of the relative errors are also shown as an indication of the magnitudes of the errors. Second, the sums of the concentrations of each component totaled for the four solutions are shown, along with the averages of the relative errors and absolute values of the relative errors. In this representation, the values found for zero concentration components,
ANALYTICAL CHEMISTRY, VOL. 56, NO. 12, OCTOBER 1984
2198
including each of the three detector phase angles corresponding to a 45' shift from the phase angles of each of the three components in combination with four sets of excitaand tionlemission wavelengths: 360/381,360/417,360/429, 3601453 nm. All possible combinations of three condition sets were then solved for the component concentrations by using eq 4. Results are shown in Table I11 for four different combinations which yielded similar results that were much better than the results for the rest of the combinations. Results are also shown for a fifth combination in which the detector phase angle was not varied (the results shown are for the best of the constant phase angle combinations). Finally, two sets of steady-state results for the same set of solutions used in the PRFS studies are shown, one using the emission wavelength maximum of each component (381,417, and 429 nm) and the other using the maxima of anthracene and POPOP (381 and 417 nm, respectively), along with 453 nm. The measurement conditions used for each of the determinations shown in Table I11 are summarized in Table IV. Precisions are shown in Table I11 for both the steady-state and the PRFS measurements. The relative precisions for the PRFS measurements are similar to the optimal precisions shown in Table I and are approximately twice as imprecise as the steady-state measurements. Accuracies are expressed in Table I11 in the same way as in Table I1 (described above). Best overall accuracies were obtained by using PRFS condition set 4, and these results show a 4-fold improvement in accuracy over the steady-state results, with a 10-fold decrease in error magnitude. All of the other PRFS condition sets show improvement over the steady-state conditions but to a lesser degree. The accuracy of the constant phase angle conditions (PRFS set 5) is as good as that of PRFS set 4 but with a greater error magnitude. It should be noted that in most of the PRFS determinations of binary mixtures (solutions 1-3), negative values are obtained for the zero concentration components, providing a convenient indication of the absence of a component. It is interesting that the best steady-state results are obtained by using the three emission maxima for the components
Table 11. Constant Wavelength PRFS Determinations" concn, pM,a t null phase angles of 207.0°, 183.8', 184.5' true value found
soln
conc, pM,a t nonnull phase angles of 148.2', 156.3', 164.4' true value found
1
A P M
5.10 0.645 0
4.46 -0.00760 1.10
6.20 0.567 0
7.43 0.521 -0.0305
A P M
5.10 0 0.610
5.49 0.0302 0.465
6.20 0 0.502
7.21 -0.732 1.01
A
0
P M
0.645 0.610
-1.81 -0.635 2.94
0 0.570 0.505
-0.190 -0.871 1.63
A P M
3.40 0.430 0.407
2.15 -0.545 2.12 23 157
4.14 0.378 0.335
4.63 0.196 0.525 13 182
2
3
4
av % error av 1% error1
total (sum of concentrations for all solutions)
A P M
13.60 1.720 1.637
10.29 -1.16 6.63 38 165
av % error av 1% error1
16.54 1.515 1.342
19.08 -0.886 3.13 -3.3 102
= 400 nm. A = anthracene, P = POPOP, M
"A,, = 355 nm, A,, = Me,POPOP.
which have undefined relative errors, can be incorporated into the summation of the total concentrations found and thereby are taken into account in the calculation of error. Phase-Resolved and Steady-State Determinations Using Wavelength Selectivity. Phase-resolved measurements were made for the series of four solutions shown in Table 111. A total of 12 measurement conditions were used,
Table 111. Phase-Resolved (PRFS) and Steady-State (SS) Determination Results Using Wavelength Selectivity"jb found true value
sol
PRFS1
PRFS2
PRFS3
PRFS4
PRFS5
SS1
SS2
1
A P M
5.200 0.645 0
5.461 0.616 0.011
5.534 0.606 0.024
5.229 0.632 0.003
5.432 0.607 0.023
5.274 0.640 -0.008
5.370 0.632 -0.011
5.325 0.640 -0.021
A P
5.200 0 0.610
5.324 -0.013 0.612
5.324 -0.013 0.612
5.351 -0.025 0.621
5.250 -0.012 0.612
5.425 -0.026 0.622
5.527 -0.023 0.626
5.168 0.040 0.548
0 0.645 0.610
-0.167 0.668 0.590
-0.151 0.666 0.593
-0.429 0.702 0.564
-0.130 0.665 0.593
-0.445 0.704 0.561
0.208 0.598 0.668
-0.452 0.713 0.524
3.467 0.430 0.407 av % error av 1% error1
3.381 0.460 0.388 0.35 3.7
3.446 0.451 0.399 0.64 3.2
3.363 0.433 0.415 0.47 3.3
3.192 0.454 0.398 -0.50 3.7
3.556 0.437 0.407 1.4 3.3
3.681 0.424 0.416 2.2 4.5
3.486 0.458 0.373 -1.6 5.9
13.867 1.720 1.627 av % error av 1% error1
14.000 1.731 1.601 1.1 1.1
14.153 1.710 1.628 0.53 0.91
13.514 1.742 1.603 -0.90 1.8
13.774 1.714 1.626 -0.43 0.43
13.800 1.755 1.582 -0.43 1.8
14.786 1.631 1.699 1.9 5.4
13.527 1.851 1.424 -2.3 7.4
0.40
0.36
0.28
0.30
0.38
0.13
0.12
2
M 3
A P
M 4
A P M
total
A P
M
measurement precision (RSD), 70
" Determination
conditions are shown in Table IV.
Results expressed as p M in cuvette.
Anal. Chem. 1984,
intensity of a particular component is usually much greater than that of other components in the samples being analyzed, the conditions can be chosen to optimize the measurement of the weaker component.
Table IV. Summary of Measurement Conditions for Determinations Shown in Table 111"
determination
hex,
nm
Aem
nm
PRFS 1 1 2 3
360 360 360
381 453 417
360 360 360
381 453 417
360 360 360
417 381 453
360 360 360
381 453 417
360 360 360
381 417 453
360 360 360
381 417 429
360 360 360
381 417 453
PRFS 2 1 2 3
PRFS 3 1 2 3
PRFS 4 1 2 3
PRFS 5 1 2 3
ss 1 1 2 3 ss 2 1 2 3
2199
56,2199-2204
CONCLUSIONS The work described here demonstrates the improved accuracy that can be obtained when wavelength selectivity is combined with fluorescence lifetime selectivity. When PRFS is used, this improvement can be achieved without increasing the number of measurements that must be taken or the complexity of data analysis. The only differences are that phase-resolution equipment is used, detector phase angles must be adjusted in addition to wavelengths, and the molar fluorescence intensities used in the simultaneous equations are phase-resolved rather than steady-state values. In all other respects the PRFS and the steady-state procedures are identical. Although this work has involved the use of three independent measurements for determination of three unknowns, the precision of the determinations could be improved by the use of more measurement condition sets to generate more equations for overdetermination of the components. For components with extensive spectral overlap, the use of selectivity based on fluorescence lifetimes is especially attractive, and the PRFS approach offers a methodologically convenient means for implementing this selectivity. Registry No. POPOP, 3073-87-8; Me2POPOP, 1806-34-4; anthracene, 120-12-7. L I T E R A T U R E CITED Veselova, T. V.; Cherkasov, A. S.; Shirokov, V. I . Opt. Spectrosc. (Eng/. Trans/.) 1970, 29, 617-618. Lakowicz, J. R.; Cherek, H. J . Blochem. Biophys. Methods 1981, 5 , 19-35. Birks, J. B.; Dyson. D. J. J . Sci. Instrum. 1961, 38, 282-285. Spencer, R. D.; Weber, G. Ann. N. Y . Acad. Sci. 1969, 158, 381-376. Lakowicz, J. R.; Cherek, H. J . Blol. Chem. 1981, 256, 6348-6353. McGown, L. B. Anal. Chlm. Acta 1984, 157, 327-332. Lakowicz, J. R.; Cherek, H.; Baiter, A. J. Blochem. Blophys. Methods 1981, 5 , 131-146. Ware, W. R.; Baidwin, B. A. J . Chem. Phys. 1964, 4 0 , 1703-1705.
Each of the three independent sets of conditions are given for each phase-resolved (PRFS) and steady-state (SS) determination. (SS l ) , whereas all of the best PRFS results are obtained by using the emission maxima for anthracene and POPOP and the wavelength (453 nm) at which the relative contribution of anthracene is low compared to POPOP and Me2POPOP. For a particular multicomponent system, the optimal combinations of wavelengths and detector phase angles must be determined experimentally and will be dictated by the particular requirements of the analysis. For example, if the
RECEIVED for review April 2, 1984. Accepted June 6, 1984.
Versatile, Efficient Raman Sampling with Fiber Optics S c o t t D. Schwab and Richard L. McCreery*
Department of Chemistry, T h e Ohio State University, Columbus, Ohio 43210 A flber optic Raman probe is described in whlch both the exclting laser ilght and the collected Raman scattering are conducted by optlcal fibers. The technique requires no alignment of sample wlth Input beam or collection optlcs, and the sample may be a great dlstance from the spectrometer or In a hostlle environment If desired. Theoretical caiculatlons demonstrate what factors in probe design determine the collection efficiency and the sampling depth of the probe. Dependlng on configuratlon, the Raman signal from the flber probe was from 1 to 9 tlmes as large as that from a conventional llqukl sampilng system. I n addttlon to hlgh collection efflclency the flber probe does not employ a focused beam, so the lncldent power denslty at the sample Is as low as one four-hundredth that of a conventional focused system. Appllcatlons to liquids, solids, low-temperature samples, and electrochemically generated species are described.
Table I. Transmission of Ensign-Bickford HC-212-T Fibers as a Function of Laser Wavelength
laser A, nm
Y,O dB/m
514.5 496.0 488.0 476.5 457.9
0.018 0.014 0.014 0.020
0.021
y is defined as (10/b) log (Zo/Zt) where I,, is intensity incident on a length b of fiber and Zt is the transmitted intensity. y is dea
termined from a ratio of transmissions through fibers of different length and is independent of coupling losses.
A variety of applications of optical fibers to spectroscopic problems have been described, with particular emphasis on UV-vis absorption and fluorescence techniques ( I ) . Fiber-
0003-2700/84/0358-2199$01S0/0@ 1984 American Chemical Society
