1730
Anal. Chem. 1988, 58, 1730-1733
Surface Analysis of Filter Papers Used in Room-Temperature Phosphorimetry M. M. Andino, M.A. Kosinski,' and J. D. Winefordner* Department of Chemistry, University of Florida, Gainesville, Florida 32611 Room-temperaturephosphorescence Is observed from compounds placed onto a SdM substrate, usually in the presence of a heavy-atom enhancer. I n order to better urderstand the surface processes, X-ray photoelectron spectroscopy studies of the surface of Whatman No. 1 filter paper are performed before and after the spotting of a iumlnescent compound and/or heavy-atom solullon onto the surface of the paper. Heavy atoms such as lodlne and thdlknn, and the compounds 3,5-dllodotyrosine, I-hydroxytryptophan,carbaryl, and bls( 8quinollnate)platinum(II)are used as probes. The observed chemical shifts and fractional areas of each class of carbon atom on the surface (as determined from computer curve fHtlng) are In reaeonable accord wHh the molecular structure of the flber. The different photoelectron peaks observed on the surface of the treated papers are identified and the elemental ratios determined. Varlatlons in the binding energies and the elemental ratlos provide information concernlng changes in the surface composition.
Room-temperature phosphorimetry (RTP) is becoming increasingly important as an analytical tool. It has proven to be a sensitive, selective, and simple method of analysis for many organic compounds, including drugs (1-3), pesticides (4-6), and compounds of environmental and biomedical concern (7). Several reviews discussing the theory and applications of RTP have been published (8, 9). Phosphorescence emission intensity depends on the microenvironment of the phosphor. The rate of intersystem crossing processes that give rise to phosphorescence is known to be enhanced by the presence of a heavy atom, which could be directly bonded to the molecule or added to the lumiphor microenvironment (10,11). In addition, in order to observe emission from the triplet excited state, the molecule must be immobilized in a rigid medium to minimize nonradiative collisional deactivation. At room temperature, phosphorescence is observed from some compounds placed onto a solid substrate. It is believed that the substrate physical and chemical interactions with the analyte and/or the heavy atom restrict radiationless deactivation (12, 13). Analysis of the surfaces from which the phosphorescence emission is observed could help us understand better the nature of the interactions and processes occurring at the surface of the support. We have selected X-ray photoelectron spectroscopy (XPS) to analyze and study the surface of the substrate with and without the presence of heavy atoms and/or luminescent molecules. The soft X-rays provide a nondestructive analytical tool that is also surface sensitive. XPS provides information about the elemental composition with a sensitivity of parts per thousand and about the chemical bonding states of the elements detected on the surface (14). Through grazing angle experiments, surface layers of different thickness can be examined (15). A pure cellulose filter paper, Whatman No. 1,was selected for the surface studies. Cellulose supports are commonly used Presently at Motorola, Inc., Ft. Lauderdale, FL 33322. 0003-2700/86/035S-1730$01.50/0
as substrates for RTP. They provide a rigid matrix for the phosphor and are also convenient, simple to use, and inexpensive. Unfortunately, there are several disadvantages associated with their use. The surface of filter papers as observed from Figure 1is very rough, irregular, and filled with interstices between cellulosic fibers. Although these factors may help protect the phosphor from oxygen, penetration of the analyte and/or of the heavy atom into the bulk of the support could be extensive. Simultaneous with penetration, there is also chromatographic migration of the molecules deposited on the surface. Both factors affect the precision, accuracy, and sensitivity of the analysis. X-ray photoelectron spectroscopy allows the qualitative and semiquantitative analysis of filter papers spotted with heavy atoms and/or phosphor solutions, provided they are present in amounts within the limit of detection of the particular element used as the probe. Elements are identified through the characteristic binding energies of their photoelectrons. The relative amounts of heavy atoms and of molecules on the surface of the paper can be determined from the photoelectron peaks. The photoelectron peaks are normalized with respect to time, the photoelectron cross sections (Scofield X-ray absorption cross sections (15)),and the instrumental factor (which is dependent on the kinetic energy of the photoelectron peaks). The extent of penetration of the analyte and of the heavy atom is observed through grazing angle experiments. There are several requirements that must be met by the lumiphors and heavy atoms to be used as probes. The compound should luminesce or specifically phosphoresce when spotted on filter paper. It should have a very low vapor pressure to withstand the vacuum conditions under which the XPS analysis is performed. The compound solubility must M in water or ethanol to allow for sufficient surface be coverage. In order to increase the sensitivity toward XPS, the presence of an element with a high photoionization cross section is desired. Salts of the heavy-atom enhancers, iodide and thallium(I), and the luminescent probe molecules, bis(8-quinolinate)platinum(II), 3,5-diiodotyrosine, 5-hydroxytryptophan, and carbaryl, were used as probes. Iodine and thallium, which are good heavy-atom enhancers commonly used in RTP, have relatively high cross sections for X-ray absorption (16). Tyrosine phosphoresces at room temperature in the presence of potassium iodide (17). A substituted tyrosine containing two iodine atoms per molecule was chosen as a probe so as to enhance its sensitivity toward XPS. The iodine photoelectron peak is used to determine the relative amount of 3,5-diiodotyrosine present at the surface of the filter paper. Both 5-hTdroxytryptophan and carbaryl are good phosphors (18,19). The phosphorescence emission intensity is increased by the addition of heavy atoms such as iodide and thallium.
EXPERIMENTAL SECTION Apparatus. The XPS studies are done with a Kratos XSAM 800 photoelectron spectrometer. The collection, presentation, and quantification of the data are done with a DS800 data system. The spectra are obtained by using nonmonochromatized Mg Kcu radiation (1253.6 eV) with a base pressure of lo-@torr and typical operating parameters of 12.5 kV and 17 mA. The instrument is operated in the FRR (fixed retarding ratio) mode. An electron 0 1986 American Chemical Society
ANALYTICAL CHEMISTRY. VOL. 58. NO. 8. JULY 1986
1731
Table 1. XPS Data for Filter Paper Whatman No. I binding element clew energy,'eV I carbon I -C.C 285.0
chemical fractional O/(CII+ shift?eV areasb CIII) 9.7 (7.6)
I
I
I1
-C-0
286.4 (288.2)
1.4 (1.7)
111
-C-0
I I
287.6 (288.6)
2.6 (3.2) 13.3 (21.7)
I
76.9 (70.9)
0
Fbuv 1. Elechon microscope photograph of Whatman No. 1 papa. The llne in the photcgraph represents 100.0 pm.
flood gun is used at all times to minimize differential charging of the filter paper. Reagents. Filter paper Whatman No. 1 is made by Whatman Laboratory Products, Inc. (Clifton. NJ). Potassium iodide and thallous nitrate are purchased from Fisher Scientific Co. (Fair Lawn, NJ). Carbaryl was donated by the Environmental Protection Agency (USA). The compounds 3.5-diiodotyrosine and Shydroxytryptophan are obtained from Nutritional Biochemicals Corp. (Cleveland. OH) and Sigma Chemcial Co. (St. Louis, MO), respectively. The bis(Bouinolinate)olatinum(II) comolex is svnthesized following the kerature pkedure (& TheI ;wen@ ).. ;sed in the preparation. Bquinolinol and potassium cetrachloroplatinate(ll), are obtained from Eastman Organic Chemicals (Rochester, N Y ) and from Strem Chemicals. Inr. (Newburyport, MA).respxtively. The solvents used are absolute ethanol (Florida Distillers Cu., Lake Alfred. FL). Nanopure deionized water (Harnstead System of Sybron. COI. and N.N.dimethylformamide tDMF, (Fisher Scientific Co., Fair Lawn. NJI. Procedure. Samples of the filter paper Whatman No. 1 are taken from the middle of the box. A ntnck of papers are perforated to produce 1-cm.dinmeter cirrles or disks. Disks from the middle of the stack are c h w n (sincethese have been less exposed to air) and used immediately to avoid contamination. The disks are mounted on the spectrometer probe with douhlesided tape. Roth treated and untreated filter paper disks are analyzed, Papers are spotted after mounting with 6 rrL of the heavy atom and/or the luminescent compound solutions by means of an SMI micropetter (Emeryville. CAI. The samples are placed in the spectrometer immediately afterward to avoid prolonged exposure to the atmusphere: they are dried lapproximately 2' hl at rmm temperature and under vacuum (lo-?torr) inside the treatment chamber of the spectrometer. The following solutions are used in the studies: 0.1.0.5, 1.0, and 2.0 M K I aqueous solutions, 0.1 M TINO, in water. 2.0 x I@ M 3,bdiiodotyrarine in ethanol. 0.01 M Shydroxytryptophan in M carbaryl in ethanol, and a saturated solution water. 5 X of the bis(8-quinolinatelplatinum(lI) complex in N.N-di. methylformamide. Potassium iodide and 3,Sdiiodotymsine pellets are made with a pellet press. Both the die plug and the die pellet receiver of the press are thoroughly cleaned and dried before use. The pelletn are mounted on the probe with double-sided tape and placed in the spertrometer for analysis.
RESULTS AND DISCUSSION XPS Studies of Whatman No. I Filter Papers. The XPS spectrum of Whatman No. 1 filter paper shows mainly C Is and 0 Is photoelectron peaks. The C Is peak is observed to consist of three main classes of carbon atoms present in wood components (21). Class 1 atom are those bonded to only carbon or hydrogen atoms. and class 11 and class 111 carbon atoms have one and two bonds with oxygen atoms, respectively. The relative abundance of each class of carbon atom on the surface was determined by computer curve fitting of the different classes of carhon peaks. The observed rhemical
oxygen 532.8 (533.5)' 0.83 'The binding energies are referend to the 285-eV hydmarbon C 1s peak. 'Values in parentheses were taken from ref 21. They are included for comparison purposes. 'Oxygen binding energy is taken from ref 22; it is the value corresponding to a McPhersan ESCA 36 soectrometer.
Qua 2. Molecular structure of ceIIubse powmer
shifta and fractional areas of these peaks are in reasonable agreement with previous reports (22). The results of the analysis of Whatman No. 1 are presented in Table I. The cellulose polymer as observed from Figure 2 contains only class I1 and I11 carbon peaks. Experimental results show that the highest contribution to the C 1s peak comes from these two classes of carhon atoms as evidenced from their fractional areas. C h I carbon atoms. as observed from Table 1, constitute 10% of the carbon atoms on this particular sample. The fractional areas of class I carbon atoms vary significantly from sample to sample. For example, C Is peaks containing up to 25% contribution from class I carbon atoms have been obtained. Although this contribution may be due to residual impurities present on the paper, it seems that the major contribution comes from contamination from the atmosphere or the spectrometer vacuum system. The average ratio of the normalized areas of oxygen to those of class I1 and I11 carbon atoms as determined from the analysis of 11 samples is 0.83 0.05. The ratio compares favorably with the theoretical value (0.83) and the literature reported value (0.79 0.04) (22). Surface Analysis of Filter Paper Treated with Heavy Atoms and/or Lumiphor Solutions. The presence of compounds containing elements other than only carbon and oxygen can be readily detected on the surface of the filter paper provided the element sensitivity toward XPS analysis is high and/or the compound surface coverage is relatively high. The binding energies of the observed photoelectron peaks are referenced to the class I (hydrocarbon-type) C 1s peak at 285.0 eV. The binding energies for the different photoelectron peaks observed from the surface analysis of the treated papers are tabulated in Table I1 The I 3d5,, binding energies as observed from the measurement of potassium iodide and 3,5-diiodotyrosine pellets are also included. The energy of an emitted core electron may be altered depending on the type of chemical bond formed by the element in question. There is a shift of 2-4 eV toward higher binding energy of the I 3d5,, photoelectron peak in 3,5-diiodotyrosine as compared to that of ionic iodine in potassium iodide as observed from the measurement of both pellets and papers treated with these compounds. The differences in binding energies reflect the differences in the chemical state of iodine in these two compounds. There is no evidence of
*
*
1732
ANALYTICAL CHEMISTRY, VOL. 58, NO. 8 , JULY 1986
Table 11. Photoelectron Binding Energies from XPS Spectra of Treated Filter Papers (and of I 3ds/2from Analysis of Pellets)
c 1s (II)*
c 1s (1II)C
0 1s
I 3d5jz
1.0 M KI (5Ie
287.3 (0.2)'
289.1 (0.2)'
533.9 (0.1)'
1.0 M KI and 0.01 M 5-hydroxytryptophan (2)e X 10" M 3,5-diiodotyrosine
286.7 (0.2)' 286.7 (0.3)'
288.4 (0.5)' 288.4 (0.4)'
533.4 (0.3)' 533.7 (0.3)'
619.7 (0.2)' [619.9 (O.6)ldJ 619.3 (0.4)' 622.1 (0.3)' 1623.6 (0.4)ldf
0.1 M TlNOs (3)' 0.1 M T1N03 and 5 X
287.3 (0.1)' 287.2 (0.1)'
289.0 (0.1)' 289.0 (0.1)'
534.0 (0.2)' 534.0 (0.2)f
spotting solution"
2
M carbaryl (2Ie
T1 4fij2
120.8 (0.6)' 121.1 (0.5)'
"Six microliters of solution is used. bClass I1 C 1s photoelectron. 'Class I11 C 1s photoelectron. dValues in brackets are the I 3d5,2 binding energy as observed on the compound's pellet. eNumber in parentheses is the number of samples analyzed. 'Number in parentheses represents the standard deviation. a strong chemical interaction of iodide ions with the cellulose fiber. Complex formation or strong chemical interaction between iodide or thallium(1) ions with the organic phosphors could alter the observed binding energies of the ions. A chemical shift of approximately 2-4 eV toward higher binding energies is expected for iodine based on the results obtained for ionic and covalently bound iodine atoms. As seen from the results in Table 11,the presence of 5-hydroxytryptophan and carbaryl do not alter significantly the binding energies of iodide and thallium(I), repectively. There is no evidence from the XPS studies of a strong chemical interaction between the heavy atom and phosphor spotted on the surface of the paper. The small shifts observed on the C 1s class I1 and class I11 of the treated papers are very small (ranging from 0.3 to 1.5 eV). The variability is within the error generated of the curve fitting process.
Table 111. Elemental Ratios from XPS Analysis of Treated Papers
Elemental Ratios from XPS Analysis of Treated Papers. The absolute amount of an element present within the depth sampled by X P S is difficult to determine; however, the
"The element's normalized peak area is ratioed to the added normalized areas of class I1 and I11 C 1s photoelectron peaks. Six microliters are used for spotting. cThestandard deviation on the iodine ratio for three samples spotted with 1.0 M KI is 0.002. The variability in the determination of the thallium ratio for two samules suotted with 0.1 M TINO, is 0.0001.
elemental ratios are readily obtained by ratioing the normalized areas of the corresponding peaks. The normalized areas of 0 Is, 13d5/2, and T1 (4f5/2 and 4f7/,) were ratioed to class I1 and I11 C 1s peaks (obtained through a peak synthesis of C 1s peak). The results are given in Table 111. The addition of 6 1L of an oxygen-containing organic molecule at M does not alter sigconcentrations of approximately nificantly the oxygen-to-carbon ratio obtained from untreated filter paper under the experimental conditions used. Iodide and thallium(1) ratios (I/C and Tl/C ratios in Table 111) from papers treated with 0.1 M KI and 0.1 M TlN03 are practically the same within the experimental error of the analysis. Because the ionic radii of I- and T1' are similar, namely 2.12 and 1.54 A, respectively (23),they are equally retained on the surface as examined by XPS, despite their charge differences. There is a linear increase of the iodide ratios with increasing concentration of iodide in solution as observed from the results in Table 111. The "heavy atom effect" of iodide on the RTP response of several phosphors is known to depend on the iodide concentration in solution (24-26). The increase in the phosphorescence intensity of several indole derivatives, as determined from previous studies (24), within the range of I- concentrations of 0.1-2.0 M correlates well with the results of the XPS studies: a linear increase in the phosphorescence intensity due to a proportional increase in the amount of Iions available at the surface of the paper. It is also important to recall that a minimum concentration of heavy atom in solution is required in order to observe the "heavy atom effect" in RTP (24). The iodine photoelectron peaks obtained from the surface of papers treated with solutions of low iodide concentration (
