1-?“ 7 xmfig- .er '. ‘ .2 'I4V, f, fIIgIs! 2d 3:331." I. f: '_p- I 1k?“ _" ' _a Vv‘ J2. 15¢:— .l 3.; A fi:» -- 3m Vii-"L; 'II-V :I i... ‘VVJ‘JVVE 7‘». lI "JUH' ‘ “I? . 31's.. _ ‘NI““-' IV. \( v ‘ A ff’g.‘ .(IIFI IIII "IV;.IVV - ..; ‘- "Jfl‘5Afl‘ I 0" IN V \ {I ‘fi—b I. |I ') I| .V‘I’N . .3: 2359: 2-1;:43‘ .1} n “V‘- ‘r: 3‘ .. Iii“ I‘quI .- 22:13“? 5“; '2':::.I..;‘.,":;4£;'i;4f TIJ: “I L . .If-fI m3 IV.IIrvI'V-'-.VVI‘I 53514 fI-V- ‘ \H-VVIIIOJ'VNIV ' I: H“ '14. .VIVI,‘VV' V I VVVVQVV'VVVVV Hg ‘. I; " :. H‘Zytu:llll ‘V‘ VV'J‘V‘ 2" l\""h't ".le III]! ‘2"'- “"u‘ I“ W“ “Ii. .‘UII ' I :I . . V . V ,VVVVVVII‘IVII-'33IVI...I:PVVIIVVVIIVV 'LIJIT‘ 3V ‘lVV ‘II'I-fIfw ‘I. V (*1 jg.‘ fil‘ ‘l‘ “m; {$166. LI} I'IVfIVV " “ I. VVV lI‘CIJI‘i-“FI'I{E‘eraq‘h}: .11“: "' 1‘1““.IVVI‘rIR' 3‘4- l‘ V. It "’""‘"'W"*I“’I' III -"" ' 7. .. I .I'IE'I'IE"~'I.I1III4=" ".5"#I'..-FII’i" " '!'I III WIT-:32“... I? ’ “5 :NVI r" "IIIII ‘V‘IIVVEIIV . V . - . it "LII“ .If$£\;“I"I '4: I. ‘1' V I I, .' I I I :‘|'v-::‘I"."‘ VV :3: ‘H‘ )‘lfl' V V V I. V Il'li MR“... .1"... IDA. VV .. VVV , VI _IV! V VVV V “I ‘I‘WLVVVW "t- I' .4‘ Wt; ‘luI .. IV . ( 'YVV'IV' V-ithIVVV‘éVI‘. VleVIVKVV: 13:02! :fVV V "Til" .‘VV' . I‘V'V‘ 43:?" . -I’ VV 3 V:.. ‘V‘ 4 . ‘-"~'I' I "'I "'I‘“ I" “ I‘I'II" 'I'II‘WI: . V V V VVVW'V II :i’B'I‘Il.IVV V V I'I'V . ’q; ‘ 3“” VV‘JV'flll II" ”$51.: V:1V\ 1H,}:V :V "4" W V:V‘:U'VE :t’I.,I‘I"I’II ‘V'III 1" xi", IVIV unfit”: ‘4 l V | ‘j If I 4 ‘l 5?: 0W '1 l' *I :‘ “V‘Il‘x' ‘ ‘| I" ‘ ‘ ‘I‘ I" I I‘M I!!!" “.‘II ”I“ WI‘ : V'n' ' V.V ‘ fia' INK” "H“ (1in V:V V VVVVVV I V V 'Vlll 4 WM ()V‘V'VV .I “1““ "J V}? ' ' j . VI‘?‘ .44"; I' ”I '33.. :4 ’VIVV‘ZRVQVV‘ WV V'VVVVVV :‘IIH‘II‘ V VI'VVI VV‘VV.V I V VV . VVV ...4. ‘ VIII‘ | V .. ' VII. :1. I ‘1 ‘ . “Fifi VVV.VV‘1VVVVl.‘ IlfculflllgI‘V," .3" il‘I‘I“ N! VV'IVW; I V VVVVV V VVVlVIV. V! 4V VVVVVV V . V VV 3‘ “Ti-.9 ‘ “w IIIIIII 94-4... "-' Q" IIIIWt : ‘ 4; I I ”x ~ ' I “ ' I" “ '5“' II ”I I L" " '. :I i 4 in ‘ _ ‘I ". 190‘?" 93““ I“. WWWI‘}, :I’HV ‘LIIIIIIH “AL": I“ 1’ ”L Q ,_ VVVI__V THEESIS This is to certify that the thesis entitled SOME EXPERIMENTAL AND THEORETICAL STUDIES OF CORRECTIONS FOR ABSORPTION INTERFERENCES IN RIGHT-ANGLE FLUORIMETRY presented by David R. Christmann has been accepted towards fulfillment of the requirements for Ph. D. degree in Chemistry M or professor Date October 6, 1980 0-7639 I. I B R A R y“ h1£i.‘];:3_n Ch?“ Aha-fl: " 7 Univ-cram y . LN kl. 4 0‘ ‘ 25‘ Per day Per it” RETURNING LIBRARY MATERIALS \ 3.." , Place in book return remove "‘"fl’ ‘ charge from circulation records 1 fif’ila‘e‘xi , y\. a“ SOME THEORETICAL AND EXPERIMENTAL STUDIES OF CORRECTIONS FOR ABSORPTION INTERFERENCES IN RIGHT-ANGLE FLUORIMETRY By David Ray Christmann A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1980 ABSTRACT SOME EXPERIMENTAL AND THEORETICAL STUDIES OF CORRECTIONS FOR ABSORPTION INTERFERENCES IN RIGHT-ANGLE FLUORIMETRY By David R. Christmann A correction factor has been developed for errors caused by absorption and reflection of the exciting and fluorescence'radiation in a dispersive right-angle fluorimeter. The correction factor is a function of the sample transmittance and cell reflectance at both the excitation and emission wavelengths and of a set of geometric window parameters characteristic of the instrument. Experiments are described which show the equation to be accurate to one percent or better up to a sample absorbance of 2.0 when the reemission of absorbed fluorescence by the sample is negligible. For samples con- taining fluorophores with highly overlapping absorption and emission bands, positive errors can occur in the corrected fluorescence at concentrations greater than about 10"5 M due to reemission phenomena. The instrumental conditions required for implementation of the absorption and reflection correction procedure include: l) that stray light be low; 2) that the bandwidths of excitation and emission be.narrow compared to the sample absorption bands; and 3) that the beams of exciting and fluorescence radiation be highly collimated. David R. Christmann These requirements will preclude the use of the correction equation in conventional filter fluorimetry. A major practical limitation of the absorption and reflection correction procedure is the need to make transmittance measurements at the excitation and emission wavelengths for each sample. To eliminate this problem, a unique fluorescence instrument has been constructed in which the fluorescence sample cell is shifted with respect to the excitation and emission Optics. From fluorescence measurements obtained at three cell positions, the sample transmit- tance values at the excitation and emission wavelengths and the cor- rected fluorescence are computed. A microcomputer is employed in the instrument to control cell positioning, data acquisition, wavelength scanning, and to transmit the acquired data to a minicomputer for permanent storage and data reduction. Automation of the cell shift procedure has resulted in a reduction in measurement time, elimina- tion of cell positioning as a source of error, and in simplification and improved accuracy of the window calibration procedure. With this instrument, a critical study of the precision and accuracy of absorption- and reflection-corrected fluorescence by the cell shift method has been carried out. Experiments are described which show the method to be accurate to two percent or better up to a sample absorbance of 2.7. Reemission and scattering phenomena must be negligible, however, and the bandwidths of excitation and emission must be narrow. Accuracy limitations are demonstrated when each of these conditions do not exist. Analytical precision has been shown theoretically and experi- mentally to be reduced by the cell shift correction procedure, but David R. Christmann only by a factor of about two. The sensitivity of the cell shift method is poorer than normal fluorimetric methods because of the high degree of excitation and emission beam collimation required. This is thought to be the major practical limitation of the correction procedure. To Dad, to Margaret, and to Mom ii ACKNOWLEDGMENTS The author gratefully acknowledges the guidance and sincere friendship of Professors Andrew Timnick and Stanley R. Crouch over the last four years. The author wishes to thank Ron Haas and Marty Rabb for their help on electronic matters, Tom Atkinson for his help with MULPLT and computer problems, and master instrument maker, Len Eisele, for his excellent work on the cell positioner. Thanks are due to many fellow members of the Crouch Group: to Rytis for keeping the 8/e running, to Charlie for always lending an ear, to Gene for his many valuable suggestions concerning this research, to Clay and Rob and Frank for their friendship, and to Jim for saving the author from the Xerox machine. The author expresses his thanks to Michigan State University, the Department of Chemistry, and Professor Frederick Horne for provid- ing financial support in the form of teaching assistantships and fellowships, to the National Science Foundation for research support, and to the American Chemical Society and Perkin Elmer Corporation for an American Chemical Society Analytical Division Fellowship. Finally, the author expresses his deepest thanks to his family for their support and for remaining so close when so far away, to Scott and Ken for their companionship on the lakes and streams, and to his wife, Margaret, for her love and patience and her capacity to put up with him. TABLE OF CONTENTS Chapter LIST OF TABLES ........................ LIST OF FIGURES ....................... CHAPTER I - INTRODUCTION ................... CHAPTER II - HISTORICAL ................... A. Problems Caused by Excessive Sample Absorption ...................... 1. Absorption Effects on Quantitative Fluorescence Measurements ............. 2. Absorption Effects on Fluorescence Spectra and Quantum Yield Measure- ments ...................... 3. Absorption Effects on Synchronous Fluorescence Spectra ............... 4. Absorption Effects on Other Fluores- cence Measurements ................ B. Conventional Methods for Dealing with Excessive Sample Absorption ............. l. Dilution of the Sample .............. 2. Use of Different Excitation and Emission Wavelengths ............... 3. Different Detection Geometries .......... 4. Special Sample Cells ............... 5. Two-Photon Excitation .............. C. Mathematical Corrections for Sample Absorption ...................... 1. Absorption Corrections for Trans- mission Geometry ................. iv Page viii ix ll 12 l3 l4 T4 15 17 18 l9 l9 Chapter CHAPTER CHAPTER A. B. 2. Absorption Corrections for Front- Surface Geometry ................. 3. Absorption Corrections for Right- angle Geometry .................. III - CORRECTION OF RIGHT-ANGLE MOLECULAR FLUORESCENCE MEASUREMENTS FOR ABSORPTION OF FLUORESCENCE RADIATION ........... Introduction ..................... Theory ........................ Cell Geometry .................. Some Initial Assumptions ............. The Attenuation by Secondary Absorption .................... 4. Absorption Effects on the Measured Fluorescence Signal ............... 5. The Correction Factor .............. Experimental 1. Instrumentation ................. 2. Reagents ..................... Results and Discussion ................ l. Measurement Conditions .............. 2. Determination of the Excitation Window Parameters ................ 3. Further Verification of the Secondary Absorption Correction ......... 4. A Limitation of the Absorption Correction Procedure ............... 5. A Comparison of Correction Factors ..................... IV - THE EFFECT OF REFLECTIONS ON RIGHT-ANGLE MOLECULAR FLUORESCENCE MEASUREMENTS .................. Introduction .................... Theory ....................... V Page 21 24 28 28 3O 30 30 32 35 35 39 39 39 39 4O 43 46 48 53 53 54 Chapter Page 1. Reflections at the Exciting Wavelength .................... 54 2. Reflections at the Emission Wavelength .................... 57 3. Determination of the Cell Reflectance .................... 59 C. Experimental ..................... 60 D. Results and Discussion ................ 6l CHAPTER V - AN AUTOMATED INSTRUMENT FOR ABSORPTION-CORRECTED FLUORESCENCE MEASUREMENTS BY THE CELL SHIFT METHOD ...... 64 A. Introduction ..................... 64 B. Theory ........................ 66 C. Instrument Design .................. 70 1. General Description ............... 7O 2. The Microcomputer ................ 72 3. Data Acquisition ................. 74 4. The Cell Positioner ............... 78 5. Alignment and Calibration ............ 83 D. Results and Discussion ................ 88 1. Determination of the Sample Transmittance .................. 88 2. Fluorimetric Determination of Aluminum ..................... 9l 3. Correction of Fluorescence Emission Spectra ................. 94 CHAPTER VI - PRECISION AND ACCURACY OF ABSORPTION- CORRECTED MOLECULAR FLUORESCENCE MEASUREMENTS BY THE CELL SHIFT METHOD ...... 99 A. Introduction ..................... 99 B. Experimental ..................... lOO vi Chapter Page 1. Instrumentation ................. lOO 2. Reagents ..................... lOl C. Results and Discussion ................ lOl l. Correction Accuracy When Reemission is Negligible .................. lOl 2. Effect of Reemission ............... l09 3. Effect of Light Scattering ............ 114 4. Effect of Spectral Bandwidth ........... l16 5. Correction Precision ............... llB CHAPTER VII - CONCLUSIONS ............. . ..... 123 A. Summary ....................... l23 8. Suggestions for Further Work ............. l28 REFERENCES .......................... l3l APPENDIX A - The IMBlOO PROM Monitor ............. 135 APPENDIX B - Microcomputer Control of the Fluorimeter ................... l53 vii Table boom A1 B1 B2 83 LIST OF TABLES Page Definition of Symbols in Equation (3.10) ....... 36 Definition of Symbols in Equation (4.1) ....... 56 Test of Reflection Corrections ............ 62 Cell Position Reproducibility ............ 84 Comparison of Sample Absorbance Estimates ...................... 90 Recovery of Aluminum in the Presence of Iron ....................... 93 Identity of Circuit Components in Figures Al-A4 .................... 140 Identity of Interface Circuit Components ...................... 157 1/0 Instructions for the Fluorimeter ......... 159 Keyboard Commands for Program DRC3 .......... 174 viii Hits Figure LIST OF FIGURES Page Geometry for right-angle fluorimetry with a square cell of internal dimensions b x b cm. Pathlength (cm) for absorption of exciting radiation at A nm = x8 (max.) and xa (min.). Pathlength (cm) for ab- sorption of fluorescence radiation at X' nm = y8 (max.) and ya (min.) .......... 31 Fluorescence of quinine sulfate at 436 nm as a function of fluorescein absorbance at 436 nm. A. Source and blank corrected fluorescence. B. Pri- mary absorption corrected fluorescence. C. Primary and secondary absorption cor- rected fluorescence. D. Results of apply- ing the correction factor (T)\.)'(OOL+GB)/2 to Curve 8 .................... 42 Fluorescence emission spectra of 10'4 M quinine sulfate (-) and 10'4 M quinine sulfate + 2.5 x 10'5 M fluorescein (*) A. Corrected fer blank, source intensity, and primary absorption. B. Corrected ix Figure Page for fluorescein emission and secondary absorption ................... 45 Fluorescence Analytical Curve of Fluorescein. (A) measured fluores- cence intensity; (I) primary absorption- corrected intensity; (9) primary and secondary absorption-corrected intensity; (—0 theoretical response ............ 47 Fluorescence analytical curve of quinine sulfate. A. Blank and source corrected fluorescence, w = 0.451, ma = 0.246. 8 8. Primary absorption corrected fluores- cence using the factor (”B'“o)£nTX/ (Txf”B-(Tx)wa. Cu Primary absorption cor- rected fluorescence using the factor (TA)' (wm‘wt31/2 ................. ' 51 Cell positions for the cell shift method. EX = excitation window, EM = emission window ..................... 67 Instrument components and their arrangement ................... 71 Block diagram of computer network ........ 73 Fluorescence S/N vs fluorescence photoanodic current. (x) + 22°C, (0) -30° C ................... 76 Figure 10 11 12 13 14 15 16 17 18 RMS dark current noise vs temperature for Hamamatsu R666 PMT at 600 V .......... Photograph of the cell positioner ......... Components of the cell positioner ......... Cell displacement in y-dimension as a function of lead screw rotation ....... Fluorescence profile of the cell in the y-dimension .................... Fluorescence profile of the cell in the x-dimension .................... Absorbance at excitation wavelength vs. fluorophore concentration. (x) calculated from F] and F2, (-) regression line cal- culated from spectrophotometric data ....... Uncorrected fluorescence emission spectra of 10'6 M rhodamine B in ethanol. (0) cell position 1, (23) cell position 2, ([1) cell position 3 ....... Uncorrected and absorption-corrected fluorescence emission spectra of 10'5 M rhodamine B in ethanol, (1) Response from cell position 1. (2) Response from cell position 2. (3) Response from cell position 3 .................. xi Page 77 79 80 82 85 87 89 95 96 Figure 19 20 21 22 Page Normalized absorption-corrected fluorescence emission spectra of rhodamine B in ethanol. (0) 10'6 M. (A) 2 x 10"5 M. (0)5 x 10'6 M. (0) 10'5 M .................... 97 Correction for absorption of exciting radiation by the fluorophore. A - Fluores- cence of quinine sulfate vs absorbance at the exciting wavelength, (I) corrected fluorescence; (A) measured fluorescence for cell positions 2 and 3; (0) measured fluorescence for cell position 1. B - RSD of corrected fluorescence vs absorbance ...... 103 Correction for absorption of exciting radiation by the sample matrix. A - Fluores- cence of a constant amount of quinine sulfate in the presence of increasing amounts of . gentisic acid, (A) corrected fl uores- cence; (I) measured fluorescence for cell positions 2 and 3; (0) measured fluorescence for cell position 1. B - RSD of corrected fluorescence vs absorbance ............ 105 Correction for absorption of exciting and fluorescence radiation. A - Fluores- cence of a constant amount of quinine xii Figure 22 23 24 25 Page sulfate in the presence of increasing amounts of f1 uorescein; (I) corrected fluorescence; (AL) measured fluorescence for cell position 1; (I) measured fluorescence for cell position 2; (0) measured fluorescence for cell position 3. B - RSD of corrected fluorescence vs absorbance ................... 107 Effect of reemission on the corrected fluorescence (A) and apparent sample absorbance at the excitation wavelength (I) when fluorescence is monitored out- side the overlap region .............. 110 Effect of reemission on the corrected fluorescence (A) and apparent sample absorbance at the excitation (I) and emission (O) wavelengths when fluores- cence is monitored in the overlap region ...................... lll Absorption-corrected fluorescence emis- sion spectra of rhodamine B in ethanol. (0)5 x 10'5 M, (A)10'5 M, (13):; x 10'5 M ...................... 113 xiii Figure 26 27 28 A1 A2 A3 A4 Page Corrected fluorescence (23) and apparent absorbance ([3), normalized to zero scattering agent, of a constant con- centration of quinine sulfate as a func- tion of increasing amounts of soluble starch ..................... ,° 115 Corrected fluorescence (23) and ap- parent absorbance at the exciting wavelength ([3) vs concentration for quinine sulfate excited with a Xe-Hg arc lamp and Corning 7-60 filter ......... 117 Normalized RSD of corrected fluores- cence as a function of absorbance. (O) (CFC/Fc)/(0Fc/F2), (---) theo- retical result from Equation (6.6) for E = 1/2 ...................... 121 Data Bus Buffering and Address Circuitry ..................... 137 PROM Board Memory and Decoding Circuitry ..................... 138 Control Panel Interrupt and Voltage Regulation Circuitry ................ 139 PROM Board Component Layout ............ 141 xiv Figure B1 82 B3 B4 85 86 B7 B8 Page I/O Instruction Decoding Circuitry ......... 155 Device Select Circuitry .............. 156 Photocurrent Amplifier Circuitry .......... 160 Sample-and-Hold and Multiplexer Circuitry ..................... 162 Analog-to-Digital Conversion Circuitry ..................... 164 Cell Position and Choooer Wheel Monitoring Circuitry ................ 166 Cell Positioner Control Circuitry. R1 = 2.2K, R2-R5 = VRl - VR4 = 1K, 01-08 = 2N6043 ................... 168 Monochromator Scanning Circuitry .......... 172 XV CHAPTER I INTRODUCTION Over the past 25 years, fluorimetric analysis has steadily gained popularity as a method for obtaining qualitative and quanti- tative information about molecules in solution. An increase in the volume of scientific literature on the subject is evidence of this (1). Basically, fluorimetry involves illuminating a sample solution with ultraviolet or visible radiation and measuring the intensity of fluores- cence radiation that is emitted. The fluorescence intensity is usually measured at a right-angle to the direction of excitation to discriminate against scattered exciting light. Several reasons for the increased use of fluorimetric analytical methods can be given. First, because only a limited number of com- pounds fluoresce, those that do can usually be determined selectively in a complex sample matrix. For samples containing several fluores- cent compounds, measurement selectivity is often greater than for molecular absorption spectrophotometry because there are two analyti- cal wavelengths to manipulate. A second and very important reason for the increased use of fluorimetric methods is that they are typically more sensitive than spectrophotometric methods by a factor of about 103. Because the measured fluorescence photon flux is usually low, very sensitive, high gain detector-amplifier combinations and photon counting tech- niques can be used. The fluorescence analytical signal is also directly proportional to the intensity of illumination. Sensitivity is, therefore, increased by the use of a more intense excitation source. Improvements in fluorescence instrumentation have also contributed greatly to the increased popularity of fluorimetric analytical methods. For many years the reproducibility of fluorimetric measurements was poor due to such instrumental distortions as the instability and spectral distribution of the excitation source intensity and the transmission characteristics of the excitation optical system. In most modern spectrofluorimeters these problems are corrected by ratioing the fluorescence signal to a reference signal that is pro- portional to the illuminating photon flux. The reference signal is obtained by employing a beam splitter directly before the sample cell to direct a known fraction of the exciting light to a quantum counter and detector. Other instrumental distortions due to the characteristics of the emission optical system and the wavelength dependence of the fluorescence detector sensitivity are eliminated by calibrating the detection system against a standard spectral source and applying corrections to the fluorescence data. This procedure is often carried out automatically in modern instruments by a micro— computer. The development of more intense excitation sources (in- cluding lasers), high-throughput holographic grating monochromators, low noise photomultiplier tubes, solid state amplifiers, and digital data acquisition systems has improved the sensitivity, precision, and accuracy of fluorescence measurements. Further developments in these areas and in applications of fiber optics, array detectors, and microcomputers will continue to expand the use of fluorimetric methods. Although the state of fluorimetric methodology has improved greatly in recent years, many problems remain to be solved. A major problem that continues to plague fluorimetric methods is the distortion of chemical information from a sample due to excessive absorption by the sample components. The errors that result can be classified as either primary absorption interferences, which involve absorption of the exciting radiation, or as secondary absorption interferences, which involve absorption of the fluorescence radiation. In the second chapter of this thesis the errors caused by primary and secondary absorption processes are discussed in detail. The experimental techniques that have been used to reduce absorption er- rors in fluorimetry are also reviewed. Because no experimental method has been a complete solution to the problems of excess sample absorp- tion, much work has been devoted to the development of mathematical corrections for absorption errors. ChapterII concludes with a review of the previous work on absorptionvcorrected fluorescence. The remaining chapters of this thesis describe several studies which were conducted with the purpose of eliminating sample absorption as a source of error in right-angle fluorimetry. Chapter III presents a detailed theory that extends the work of previous investigators (2) in explaining the effects of both primary and secondary absorption processes on right-angle fluorescence measurements. A correction factor for secondary absorption interferences is derived and tested, and the conditions necessary for its implementation are identified. The theory is extended in ChapterIV’to include the effects of re- flections within the sample cell at the excitation and emission wave- lengths. Correction factors for reflection effects are derived and shown to improve the accuracy of the absorption correction procedure. These treatments form the basis for the work described in the follow- ing chapters. In Chapter V, the design of a unique absorption-corrected right- angle spectrofluorimeter is described. A microcompUter is employed to shift the fluorescence sample cell so that the effective path- lengths of excitation and emission are independently varied. From fluorescence measurements at three cell positions, corrections for both primary and secondary absorption interferences are computed. Direct measurement of the sample transmittance is not required. The theory of the cell shift method is explained, and the performance of the instrument is evaluated. In the next chapter, a critical study of the precision and ac- curacy of absorption-corrected fluorescence measurements by the cell shift method is described. Accuracy limitations of the correction procedure are demonstrated when wide spectral bandwidths are involved and also when scattering and reemission phenomena occur. When these conditions are absent, the method is shown to be highly accurate. Analytical precision is reduced slightly by the absorption correction procedure, but the effect is generally insignificant. From these findings it is clear that, although limitations exist, in many fluorimetric analytical methods interferences due to all types of sample absorption can be totally eliminated. Finally, the significance of this work is summarized, and some recommendations for future work in the area of absorption-corrected right-angle fluorimetry are made. It is hoped that this work will serve to promote the use of absorption correction procedures and to improve the accuracy of many fluorimetric analytical methods. CHAPTER II HISTORICAL A. Problems Caused by Excessive Sample Absorption In this section, absorption interferences in molecular fluores- cence spectrometry are discussed according to their effects on quantitative fluorescence measurements and on fluorescence spectral data. The discussion initially focuses on right-angle detection geometry, but other detection methods are also considered. 1. Absorption Effects on Quantitative Fluorescence Measurements Primary absorption by the fluorescence analyte is a necessary process in all fluorimetric methods. It can also be the cause of analytical error. The most obvious effect of primary absorption on quantitative fluorescence measurements is that the measured fluores- cence intensity is not a linear function of the fluorophore concentra- tion. This fact is usually emphasized in fluorescence monographs (3,4) and textbooks of analytical chemistry (5,6) by the equation F = K¢10(1-10‘€bc) (2.1) where F is the intensity of fluorescence emission, K is a constant that accounts for instrumental factors, o is the fluorescence quantum efficiency, and I0 is the intensity of monochromatic illumination. The quantities a, b, and c have their usual Beer's law meanings. As the fluorophore concentration approaches zero, Equation 2.1 approaches the linear form F = 2.303K Iocbc (2.2) Because fluorescence measurements can be made at very low concentra- tions, therefore, the relationship between fluorescence intensity and concentration is usually assumed to be linear. However, some degree of error is always introduced by this assumption (7). When the primary absorbance of the fluorophore exceeds a value of 0.01, the fluorescence analytical curve begins to bend noticeably toward the concentration axis. At higher fluorophore concentrations the curve passes through a maximum and the fluorescence response begins to decrease (8). Holland (9) has explained this behavior as the result of a shift of the most intensely fluorescing region of sample solution toward the excitation face of the cell as the penetra- tion depth of the exciting radiation is reduced by increased absorp- tion. Because the emission window'must be chosen so that the cell walls are masked to prevent distortions from reflection and scatter- ing phenomena, at high enough analyte concentrations not all of the radiation from the most strongly emitting region of solution passes through the geometrical detection window of the fluorimeter, and fewer fluorescence photons are observed. Ohnesorge (10) has shown that the emission window dimensions and location affect the curva- ture of the fluorescence response. If fluorescence is monitored on the short wavelength side of an emission band, absorption of the fluorescence radiation by the fluorophore due to the overlap of its absorption and emission bands, i.e., self-absorption, will also cause the fluorescence analytical curve to bend through a maximum at higher concentrations. This occurs because the fraction of light transmitted by the solution at the emission wavelength decreases exponentially as the fluorophore concen- tration is increased (11). Monitoring fluorescence in this spectral region is not a recommended practice, but may be necessary to avoid interferences from other chromophores or fluorophores in the sample. As a consequence of these effects, a nonlinear analytical curve must be used or the samples must be diluted to make use of the linear portion of the curve at lower concentrations. Both of these alterna- tives are inconvenient and involve possible sources of error. It would be most desirable if fluorescence measurements could be cor- rected to zero sample absorption so that a linear calibration curve would result. Other absorption problems can be encountered in quantitative fluorimetry if the sample contains primary or secondary absorbing components other than the fluorOphore. If the calibration standards are not prepared in this same matrix, the attenuation of the sample fluorescence by primary or secondary absorption is not compensated, and a serious negative error in the analysis can result. Hemoglobin is known to interfere in this manner with the determination of zinc protoporphyrin in blood (12) and in the determination of serum cal- cium (l3). Often the sample matrix cannot be exactly duplicated and may vary between samples. Separation procedures prior to the fluori- metric analysis may be necessary to avoid the error. This has the disadvantages of increasing the amount of sample treatment, increasing the analysis time, and introducing an additional source of error. 2. Absorption Effects on Fluorescence Spectra and Quantum Yield Measurements It is well known that instrumental factors such as the spectral energy distribution of the excitation source and the spectral sensi- tivity of the detection system can severely distort fluorescence excitation and emission spectra unless corrections for their effects are made (14-21). Parker and Rees (14) have shown that primary and secondary absorption processes can cause spectral distortions which are just as serious. Theoretically, the fluorescence excitation spectrum of a compound should be identical to its absorption spectrum (3,22). To observe this for a given sample, however, the measured fluorescence intensity must vary linearly with the sample absorbance. As discussed above, this behavior is approached accurately only when the primary ab- sorbance of the fluorophore is 0.01 or less. When the absorbance exceeds this limit the fluorescence excitation spectrum sags below the absorption spectrum (23). The problem is compounded if the ab- sorbance of the sample is higher than 0.01. The shift of the fluores- cing region of solution out of the detection window of the 10 instrument can cause a further decrease of fluorescence intensity and possibly a minimum in the excitation spectrum in the spectral region of an absorption peak. Clearly, under conditions in which the sample primary absorbance exceeds 0.01 the true fluorescence excitation spectrum will not be observed. Primary absorption processes also affect fluorescence emission spectra. The decrease in emission intensity caused by primary ab- sorption affects all emission wavelengths so that the entire emission spectrum is attenuated by a constant factor. Parker and Rees (21) have noted that this will cause a negative error in the relative quantum yield determined from the area under the spectrum. They sug- gest: l) keeping the primary absorbance of the sample below 0.02 to reduce the error to four percent or less, or 2) matching the ab- sorbance of the quantum yield standard to that of the sample. Secondary absorption by the fluorOphore causes a loss of in- tensity on the short wavelength end of fluorescence emission spectra (21). Because of this, a serious negative error, which Demas and Crosby (24) have called the "reabsorption error", will exist in the quantum yield determined from such a spectrum. Similarly, secondary absorption by other components of the sample will lead to decreased intensity and possibly minima in the observed emission spectrum in the spectral regions where they absorb. This also causes the measured quantum yield to be low and precludes accurate quantum yield measure- ments on all but pure fluorophore solutions. 11 3. Absorption Effects on Synchronous Fluorescence Spectra In recent years, a new modification of the fluorescence method called synchronous fluorescence spectrometry has come into use (25). In this method, the excitation and emission wavelengths are scanned in tandem with a constant wavelength difference between them. The result is a unique fluorescence emission spectrum that is often a better fingerprint of a compound than its normal emission spectrum. The qualitative power of the synchronous method has been demonstrated by its use to identify crude oil samples (26) and traces of rubber (27) as well as its use in a wide range of forensic applications (28-30). Vo-Dinh (31) was able to resolve completely and identify as many as five polycyclic aromatic hydrocarbons in the synchronous spectrum of a mixture for which the normal emission spectral components were hopelessly overlapped. Theories of the synchronous emission spectrum have been published by Lloyd and Evett (32) and by Vo-Dinh (31). From these discussions, it is clear that a synchronous signal is observed only in the spectral region in which the normal fluorescence excitation and emission spectra overlap. A combination of the extent of overlap, the features of the overlapping spectra, and the spectral position of the overlap make the synchronous spectrum unique for each compound. However, these same factors make the synchronous spectrum highly subject to pri- mary and secondary absorption interferences. Because the synchronous signal is derived from the region of spectral overlap, secondary absorption by the fluorophore will 12 inevitably attenuate the signal and distort the observed spectrum. This fact has been the subject of a recent controversy about the usefulness of the synchronous method (33-35). Since the excitation and emission processes in the overlap region are not as efficient, higher than normal sample concentrations must be used to obtain detectable signals. This step increases secondary absorption inter- ference and may cause primary absorption by the fluorophore to become significant. The synchronous method appears to be useful on complex samples for which the normal emission spectral components overlap. But in these types of solutions, primary and secondary absorption by the other sample components attenuate the synchronous fluorescence signal in the same way that they affect the normal emission spectrum (33). The synchronous emission spectrum may be a valuable qualitative tool, but it is not likely to be of significant quantitative value unless the primary and secondary absorption interferences can be cor- rected. No quantitative applications of synchronous fluorescence measurements are presently known, and indeed, the data of White (36) suggest that the relationship between the synchronous fluorescence signal and fluorophore concentration is not linear. 4. Absorption Effects on Other Fluorescence Measurements Molecular fluorescence spectrometry is often used to determine chemical equilibrium constants and related thermodynamic quantities such as AG, AH, and AS for both ground state and excited state molecules (37). It is assumed in these studies that the fluorescence 13 signal is linearly related to the fluorophore concentration. Lin- earity is not always verified with a calibration curve, however. When a calibration curve is used it may not compensate for absorption interferences due to the sample matrix if the matrix is not dupli- ' cated for the calibration standards. Accurate experimental results can be very difficult to obtain. Fluorescence detection is also used to monitor rates of reaction from which rate laws, rate constants, and activation parameters are detennined. Here also, it is fundamentally assumed that the fluores- cence signal is a linear function of the analyte concentration. Primary and secondary absorption interferences can cause the signal- concentration relationship to be far from linear, and the degree of interference can change as the reaction proceeds. Thus, it could be difficult to extract a meaningful rate constant from the observed rate behavior,and all calculated results could unknowingly be in error . B. Conventional Methods for Dealing with Excessive Sample Absppption Many investigators have found ways to overcome absorption inter- ferences in particular experimental situations by altering their experimental procedure or fluorescence instrumentation. These methods and their limitations are discussed in this section and are compared to normal right-angle fluorimetric methods when appropriate. 14 l. Dilution of the Sample The most widely recommended and used procedure for dealing with primary and secondary absorption interferences is to dilute the sample until the magnitude of the absorption effect is reduced to an acceptable level (3,4). In many situations this is a simple and practical remedy. It will not always be satisfactory, however. Dilu- tion of the fluorophore almost always involves a loss of sensitivity. It is possible that to reduce the sample absorbance to an acceptable level the fluorophore must be diluted to a concentration that is dif- ficult to detect. For example, for a sample containing a weak fluoro- phore and a strong primary or secondary absorbing chromophore. the fluorescence signal might be near the detection limit and dilution would not be possible. Other disadvantages of sample dilution procedures are that they increase the amount of sample treatment and the analysis time and can be a source of error. Although it should be possible to keep dilution errors small, it is best to avoid them altogether. Finally, dilution can shift solution equilibria. The use of dilution procedures in some types of studies and analyses may be impossible for this reason . 2. Use of Different Excitation and Emission Wavelengths With modern spectrofluorimeters it is often possible to choose an excitation and emission wavelength at which primary and secondary absorption interferences are negligible. Although this is usually 15 a convenient remedy, with complicated samples it may not be pos- sible to avoid the absorption interference of one chromophore without encountering absorption interferences from other chromophores or spectral interferences from other fluorophores. Another limitation of this procedure is that sensitivity is usually lost by moving the excitation or emission wavelengths away from the fluorescence excita- tion or emission maximum. 3. Different Detection Geometries Another common method of dealing with absorption interferences is to use a different detection geometry, of which there are two types: "in-line" or "transmission" geometry and "front-surface" geometry. a. Transmission Geometry - In the transmission geometry arrange- ment, the fluorescence detector views the cell face through which the exciting light leaves the cell. The advantage of this geometry is that primary absorption interferences are less severe. As the pri- mary absorbance of the sample is increased, the fluorescing region of the solution moves toward the excitation face of the cell but always remains in the detection window. The curvature of the fluores- cence analytical curve and the resulting errors are, therefore, reduced. There are also several disadvantages of transmission detection geometry. First, primary absorption interferences are only reduced and not eliminated. 0f more importance, secondary absorption 16 interferences are usually more severe than with right-angle geometry (38). This is due to the fact that some of the fluorescence light must travel the entire width of the cell through the sample to reach the detector. Secondary absorption interferences can be even worse if the primary absorbance of the sample is high. As the fluorescing region of solution moves toward the excitation face,the average path- length through the sample for the fluorescence emission is increased and can approach the full width of the cell. Distortions of fluores- cence excitation spectra recorded with transmission geometry have been reported and blamed on this effect (39). b. Front-surface Geometry - "Front-surface“ or "reflection" geometry involves observing fluorescence emission through the front or excitation face of the sample cell. It is often used to avoid absorption interferences in fluorescence emission spectra. For very highly absorbing solutions, the exciting light is almost completely absorbed at the cell wall-solution interface. The depth of penetra- tion of the exciting radiation into the solution decreases with in- creasing analyte concentration. The path for fluorescence radiation out of the solution is, therefore, very short. Also, the fluorescing region of solution always remains in the instrument detection window. These conditions help to reduce primary and secondary absorption errors. To maintain these conditions, however, highly concentrated samples must always be used, and because of this,secondary absorption and quenching errors can occur (24). McDonald and Selinger (39) have shown that large secondary absorption errors, similar to those which 17 plague right-angle measurements, can exist in fluorescence emission spectra obtained with the front surface arrangement. Berlman (40) and Parker (3) have suggested that the fluorophore be excited at its most intense absorption peak to reduce the concentration of fluoro- phore required. This does not eliminate the interference but does make it less serious than in right-angle measurements (38). Another limitation of front-surface fluorescence detection geometry that has been noted (24,38) is that the emission spectra recorded are more subject to distortion due to the reemission of absorbed fluorescence by the sample. Quantum yields up to 50 percent greater than the true value have been measured for 9,10-diphenyl- anthracene because of this effect (41,42). 4. Special Sample Cells Several variations of the fluorescence sample cell that take ad- vantage of shorter pathlengths for the exciting and fluorescence radia- tion into and out of the sample have been shown to be effective for reducing absorption interferences. Chen and Hayes (43) were able to extend the linear portion of the fluorescence calibration curve of NADPH by almost an order of magnitude in concentration by placing a square cuvette eccentrically in the sample compartment of an Aminco- Boman spectrofluorimeter. Similar improvements were obtained when a 3 mm microcell was used in place of a normal 1 cm cuvette. With both arrangements, however, the sample absorption interferences were only reduced and not eliminated. Another problem is that interferences from light scattered and fluoresced by the cell walls is greater than 18 for the normal right-angle arrangement, especially with the micro- cell. Mitchell, Garden, and Aldous (44) described a fluorimetric cell in which a single fiber optic bundle carries the exciting light to and the fluorescence light from the sample solution. This also reduces the pathlengths for absorption. Compared to a right-angle fluorimeter, they were able to reduce the measurement error on a solu- tion with a primary absorbance of 0.25 from 63 percent to 21 percent with this cell. Although this is a significant improvement, more improvement is desirable. It is doubtful that this fiber optic design will be of significant use for dealing with absorption interferences in fluorimetry. 5. Two-Photon Excitation Recently, Wirth and Lytle (45) used the method of two-photon excitation to overcome primary absorption interferences in optically dense solutions. In this mode of excitation the fluorophore must simultaneously absorb two photons which are of half the energy re— quired for single-photon absorption. Two-photon absorption does not follow the Beer-Lambert absorption law and is characterized by a very low absorption cross section. These facts make fluorimetry with two- photon excitation essentially immune to errors due to primary ab- sorption by the fluorophore. In one experiment, a constant fluores- cence signal was measured for a constant amount of p-terphenyl as the primary absorbance of the solutions for single-photon absorption was increased to 28. 19 Although these results are promising, several limitations exist. The selection rules for two-photon absorption differ from those for single photon absorption making it difficult to reach some excited energy states. The instrumentation required for two-photon excitation is also more complicated and expensive. A laser excitation source must be used. Also, the method is not immune to secondary absorption interferences. Focusing the exciting laser beam close to the emission face of the sample cell will reduce secondary absorption errors but not eliminate them. Finally, single-photon absorption of the exciting laser light by the sample matrix will cause primary absorption errors. For complicated samples the method may not offer any advantages over conventional fluorimetry. C. Mathematical Corrections for Sample Absorption It is evident from the previous discussion that there is no universal experimental method for obtaining absorption-free fluores- cence information from a sample. For this reason, many attempts have been made to describe primary and secondary absorption interferences mathematically so that fluorescence measurements can be corrected for their effects. These efforts are reviewed here according to the detection geometries used. 1. Absorption Corrections for Transmission Geometry Probably the first attempt to describe the effects of absorption on transmission type fluorescence measurements mathematically was 20 made by Weber in 1930 (46). Two equations were proposed to describe the fluorescence intensity as a function of the primary absorbance of the sample. Measured fluorescence intensities from a transmission type filter fluorimeter were compared with values calculated from the proposed relationships, but without much success. More recent work has shown that the equations were too simple to describe the experimental results accurately. In 1938, Sen-Gupta and Ghosh (47,48) derived equations to des- cribe transmission as well as front-surface and right-angle fluores- cence measurements in terms of both the primary and secondary ab- sorbances of the sample. Although these equations were more carefully conceived, the fact that radiation absorbed by the sample matrix will not contribute to the observed fluorescence intensity was neglected. No further work based on this study is known. However, in 1944, Rollefson and Dodgen (49) independently derived an equivalent expres- sion. Using measured solution absorbances at the excitation and emis- sion wavelengths, they obtained good agreement between calculated and observed fluorescence intensities from solutions of fluorescein and of acridone. The correction appeared to be accurate for solutions of a single fluorophore, but was not tested with other primary absorbing components present in the sample. In 1973, van Slageren and workers (50) published a third inde- pendent derivation of the same absorption expressions for transmission type fluorescence measurements. Theoretical plots of the fluores- cence response as a function of sample absorbance were presented, but no experimental verification was attempted or has been described 21 since the work was published. The authors did clearly state that for the expressions to be accurate, the excitation light must be homogeneous, monochromatic, and collimated. The fluorescence light reaching the detector must also be monochromatic and collimated. It appears that a good theoretical description of the effects of primary absorption by the fluorophore on transmission type fluores- cence measurements is available. It is not clear, however, that the equations are accurate when a second absorbing component is present in the sample. No applications of the correction equations have been reported, probably because of the more widespread use of other detec- tion geometries. 2. Absorption Corrections for Front-Surface Geometry Many papers on absorption-corrected fluorescence have dealt with front-surface detection geometry. The first work of this sort was by Sen-Gupta and Ghosh (47,48) who derived an expression for the fluorescence intensity in terms of the molar absorptivities of the sample at the excitation and emission wavelengths. An equation of the same functional form was published by van Slageren g; 11. (50). In 1951, F6rster (51) derived an expression of similar form but not identical to these equations. In all three cases, no experi- mental verification of the proposed relationships was described. In 1956, Budo and Ketskemety (52,53) initiated a series of papers (54,55) that described a complicated correction for primary and secondary absorption and also reemission errors in front-surface 22 fluorimetric measurements. Their equation was similar to that of Sen-Gupta (48) but included a factor (1 + K) to compensate for re- emission of absorbed fluorescence. The factor K was defined to be the ratio of the reemitted light to that of the direct fluorescence light and was calculated by evaluating a complex integral. These workers maintained that by replacing the factor (1 + K) with a factor l/(l - K) multiple reabsorption and reemission effects could be corrected. They offered some experimental evidence as proof (55). The results seem to substantiate the accuracy of the correction procedure, but the conditions under which the correction can be used are unclear. The calculation of the factor K was also very complicated and not explained well. In 1961, Melhuish (41) modified the equations of Budo £3 11. (52) to include a correction for the movement of the fluorescing region of solution away from the front face of the sample cell as the sample is diluted. The limiting forms of the equation at high and low primary absorbance were discussed, and the fermer was applied to cor- rect Stern-Volmer plots for anthracene and perylene to linearity. More recently, a point by point correction of the emission spectrum of perylene has been published (56). The remainder of the literature on absorption-corrected fluores- cence for front-surface geometry has focused on correcting fluores- cence quantum yields for absorption interferences. Birks (38) des- cribed a simple correction which involves multiplying the observed quantum yield o by a factor l/(l - a), where a is a quantity called the "self-absorption probability". This factor was determined from 23 the integrated emission spectrum both in the presence and in the absence of self-absorption phenomena. The correction was later ex- tended to cover interference due to reemission of absorbed fluores- cence by letting the correction factor be l/(l - a + a6). This cor- rection factor was claimed to be valid for all observation geometries except the front-surface case when ¢ exceeds the value of 0.3 (57). If p is greater than 0.3, the difference between the true and observed . quantum yields is great enough to make the correction inaccurate. For the correction to be exact, the true quantum yield would have to be used to calculate the correction factor. A quite different approach was taken by Rohatgi and Singhal (58) in 1962. These authors described the "average molar absorptivity for self-absorption" which was derived from a simple exponential attenua- tion model of secondary absorption interference. They showed how this parameter could be determined experimentally, and later derived a relatively simple, new expression to correct the observed quantum yield from measured values of the parameter (59). A linear Stern- Volmer plot was presented to verify the reliability of the method. The simplicity of the calculations was stressed. Finally, Alan Mode and Sisson (60) described a correction scheme in which an integral was numerically evaluated to obtain a correction factor for primary absorption interference in front-surface quantum yield measurements. They were successful in obtaining constant cor- rected quantum yields from quinine sulfate solutions in which the primary absorbance varied from 2 to 24. This required repeating the numerical evaluation for each value of the sample absorbance. 24 Reemission and secondary absorption interferences, which others have found to be significant, were not considered in their treatment. 3. Absorption Corrections for Right-Angle Geometry Many of the first attempts to describe the effects of primary and secondary absorption on right-angle fluorescence measurements were inaccurate because of a failure to account for absorption by sample components other than the fluorophore and because of geometrical simplifications that do not represent the true experimental arrange- ment. The equations given by Sen-Gupta (47) and by F6rster (51) are of doubtful validity because they were derived from the assumption that the observed fluorescence originates from a point source. This is a poor approximation of the conditions in a real right-angle fluorimeter. Braunsberg and Osborn (61) and Lauer (62) neglected the excitation and emission window dimensions, which are usually chosen to mask the cell walls. Henderson (11) considered the experimental geometry of the fluorescence emission window more carefully, but ignored the finite width of the excitation beam. He found good agree- ment between calculated fluorescence values and a measured analytical curve for biacetyl and claimed this to be satisfactory verification of his equation. The experiment did show that his consideration of primary absorption interference by the fluorophore was probably correct, but did not test the expression for primary absorption by other sample components. The test also did not extend to high secondary absorbance values at which the geometrical simplification 25 that was made might create errors. Thomas and coworkers (63) have presented a more complete con— sideration of the measurement geometry in deriving an equation for the right-angle fluorescence intensity. Several calculated and measured analytical curves were presented which agreed well below a primary ab- sorbance of about 0.5, but showed large discrepancies at higher ab- sorbances. The differences between the observed and calculated sig- nals were attributed to inaccurate absorption measurements on the more concentrated solutions. The equation was claimed to be accurate, but as in Henderson's work (11), the equation considered only primary ab- sorption by the fluorophore and was not tested under conditions in which the secondary absorbance of the sample was significant. St. John g; pl. (64) included expressions for the effects of primary and secondary absorption in a theory of signal-to-noise ratio in luminescence spectrometry. Their equations were adapted in part from a theory of self-absorption in flame photometry (65). It is not clear that this extension of the theory is valid, however. Leese and Wehry (66) have presented derivations of correction factors for primary and secondary absorption errors in determinations of Stern-Volmer quenching constants by right-angle and front-surface fluorimetry. Although their data support the validity of their method, the correction factors involve complex integrals that must be evaluated numerically. The equations are also only applicable to Stern-Volmer quenching data. This work is clearly not a general solution to the absorption interference problem in right-angle fluorimetry. 26 Other workers have made more significant contributions to absorp- tion corrected right-angle fluorimetry. In 1957, Parker and Barnes (23) proposed a unique primary absorption correction factor that is a function of only the emission window dimensions and the absorbance of the sample at the excitation wavelength. The factor was claimed to be valid for not only primary absorption by the fluorophore, but also for primary absorption by the sample matrix. The theoretical origin of the expression and the experimental conditions under which it would be valid were not explained. Gill (67) attempted to answer these questions but was able only to justify Parker's equation partially. More recently, an accurate explanation of the theoretical basis of the primary absorption correction factor, experimental evidence of its validity, and the experimental conditions necessary for its implemen- tation have been published (2,68). Van Slageren §t_al, (50) have described the attenuation of right- angle fluorescence intensity by primary and secondary absorption in a manner that also accounts for reflections within the sample cell. Their equations appear to account completely for the measurement geometry and for all primary and secondary absorbing components in the sample. The conditions under which the equations apply were clearly stated and closely approximate those found in most modern spectrofluorimeters. Unfortunately, no experimental verification of their theory was attempted. Based on this work, however, Novak (69) recently derived and tested expressions to correct right-angle fluores- cence measurements for all absorption interferences without requiring the direct measurement of the sample absorbanCe. Two procedures were 27 described by which the sample absorbance or transmittance at the excita- tion and emission wavelengths can be computed from only fluorescence measurements. One procedure involves measuring the fluorescence with a second sample cell in the excitation or emission light path. The second procedure, called the cell shift method, involves shifting the sample cell with respect to the instrument window optics and measuring the fluorescence at three cell positions. The absorbance values com- puted from these measurements were used to compute a correction factor fOr total sample absorption. Some evidence for the accuracy of these correction procedures was presented, but a critical evaluation of the method has not been published. This is the subject of later chapters in this thesis. CHAPTER III CORRECTION OF RIGHT-ANGLE MOLECULAR FLUORESCENCE MEASUREMENTS FOR ABSORPTION OF FLUORESCENCE RADIATION A. Introduction In previous work by Holland £3 31. (2) and Kelly (68), a correc- tion factor for primary absorption interferences in right-angle fluorimetry similar to that presented and applied by Parker and Barnes (23) was derived and tested. It was demonstrated that right- angle molecular fluorescence measurements can be accurately corrected for primary absorption interferences when the sample absorbance at the excitation wavelength is as high as 2.0. However, the usefulness of these corrected measurements is limited unless a correction for secon- dary absorption interference can also be made. The primary absorption- corrected fluorescence intensity is directly proportional to the fluorophore concentration, the desired result, only if the sample is completely transparent to the emitted fluorescence. Many real samples contain components that absorb the analyte fluorescence. The precision and accuracy of the primary absorption-corrected measurement can be seriously degraded as a result. As discussed in the previous chapter, a correction procedure for secondary absorption interference in front-surface fluorimetry 28 29 has been developed and verified by several investigators, but the literature is not consistent on a correction procedure for right- angle detection geometry. Several different mathematical descriptions of the interference have been published for the right-angle case, but no experimental support is given for some of these models. Where experimental results are reported, the accuracy of the model is not clearly demonstrated. A major matter of dispute is whether the fluorescing volume of solution viewed by the detector can be treated as a point source of light (11). This assumption leads to a simplified correction scheme (66), but one of questionable accuracy for a real fluorimeter. Only for the cell shift method proposed by Novak (69) is clear experimental evidence given of the accuracy of a secondary absorption correction procedure. To clear up the confusion that exists in the literature, and to establish a basis for an investigation of the cell shift method, a detailed study of secondary absorption interferences in right-angle fluorimetry was undertaken. In this chapter, a general theory of secondary absorption interference for a right-angle fluorimeter equipped with a square fluorescence cell is described which incorporates the primary absorption theory of Holland g;_§l, (2). For the specific case of a dispersive instrument, a correction factor for secondary absorption interference is derived, and experimental results are presented to verify its effectiveness. The superiority of this method is demonstrated by comparing experimental results with those obtained by using a correction factor consistent with the point source assumption. 30 B. Theory 1. Cell Geometry Figure l is a top view representation of the cell geometry that is typically used in right-angle fluorimetry. The fluorescence cell is square with sides of length b cm. The thickness of the cell walls is assumed to be negligible for simplicity. The excitation beam enters the cell through the excitation window which is defined by baffle edges at distances yB and ya cm from the plane of the emission face of the cell. Similarly, the emission beam leaves the cell through the emission window which is defined by baffle edges at distances x and xa cm from the plane of the excitation face. 8 2. Some Initial Assumptions In addition to the cell geometry shown in Figure 1, it is initially convenient to make the following assumptions. (a) The excitation beam is homogeneous, collimated, and mono- chromatic with a wavelength X nm. (b) The emission beam is collimated and monochromatic with a wavelength X' nm. (c) Fluorescence photons which are absorbed in the cell are not remitted by the sample. (d) Scattered light, refractive index effects, and reflections within the cell are negligible. (e) The sample is homogeneous and contains a single fluorophore although other chromophores may be present. 31 ‘< b s» 7 1 i b A // +7 1 191“ ( a: v ll ._._mzm.rz_ muzwummeDJm _ no 5 4o«— 30:— 20“- IO:— 6 OO 5 WAVELENGTH. 400 nm Figure 3 44 Figure 3. Fluorescence emission spectra of 10’4 M quinine sulfate (-) and 10'4 M quinine sulfate + 2.5 x 10"5 M fluorescein (*) A. Corrected for blank, source intensity, and primary ab- sorption. B. Corrected for fluorescein emission and secondary absorption. 45 1+ no 7 _ o .o >._..mzm.rz_ muzwummeDJn. 50.— 4o~— 20.— IO "" 500 WAVELENGTH,nrn 400 Figure 3 46 detailed by the decreased intensity in the mixture spectrum at wave- lengths smaller than 490 nm. At wavelengths greater than 490 nm, the increase in intensity is due to fluorescein emission. Figure 3B shows the mixture spectrum after the corrected emission of the fluores- cein solution was subtracted and the secondary absorption correction applied. The spectrum of the quinine sulfate solution is repeated in Figure 33 for comparison. The accuracy with which the quinine sulfate emission is restored at wavelengths less than 490 nm clearly illustrates the validity of the secondary absorption correction. 4. A Limitation of the Absorption Correction Procedure At wavelengths larger than 490 nm in Figure 38, a residual com- ponent of fluorescein emission can be distinguished which is due to the excitation of fluorescein by the quinine sulfate emission. This is a violation of assumption 3 and illustrates the failure of the absorption correction procedure when the absorbed fluorescence is re- emitted by the sample. Reemission appears to be a potentially serious limitation of the absorption correction procedure because many fluoro- phores have highly overlapping absorption and emission spectra, and therefore, absorb and reemit their own fluorescence. The effect of reemission is shown more clearly in Figure 4 which shows the absorption- corrected fluorescence analytical curve of fluorescein in 0.1 N NaOH measured at 510 nm and excited at 313 nm. The straight line in Figure 4 is the corrected fluorescence response extrapolated from concentrations below 6 x 10'6 M for which the sample absorbance at the excitation and emission wavelengths was less than 0.05. Clearly Figure 4. 47 700 — 1 . 600—- § 2 -. “5500-— In. ‘t -- E ‘3' B 8 a d 0 ::}::::::::}::::3l 0 5 10 15 20 FLUORESCEIN CONCENTRATTON, M x 105 Fluorescence Analytical Curve of Fluorescein. (A) Measured fluorescence intensity. (I) Primary absorption-corrected intensity. (0) Primary and secondary absorption-corrected intensity. (——) Theoretical response. 48 the presence of a reemission component in the measured fluorescence signal causes an overcorrection for the effects of primary and secon- dary absorption. The error increases in magnitude as the fluorophore concentration is increased. An identical effect is observed when fluorescein emission is measured at 550 nm where only primary absorp- tion interference is encountered. At both emission wavelengths, 5Mand however, the error is insignificant at concentrations of 10' below. Reemission errors have not been studied previously for right- angle detection geometry, and no method to correct for them is of- fered in this work. Because of the complexity of the reemission error corrections that have been developed for front-surface fluorimetry (41), it is doubtful that reemission corrections for right-angle fluorimetry could be accomplished in a manner simple enough to be of practical value. Instead, the effects of reemission have been identified to characterize the practical limitations of the absorp- tion correction procedure. A further discussion of reemission errors is given in ChapterVI. 5. A Comparison of Correction Factors Both primary and secondary absorption errors have been described in previous reports as simple exponential attenuations over an average pathlength through the sample (11,66). This description implies that the fluorescing volume of solution viewed by the detector can be treated as a point source emitter located at the center of the volume element. The secondary absorption correction factor under 49 this assumption is (Tx.)'(GG+OB)/2. Figure 2, Curve 0 shows the result of applying this correction to the data of Curve B. Clearly, at absorbances less than 1 this correction compares well with Equa— tion (3.13), but at higher absorbances a positive error is intro- duced which becomes significant (>10%) above an absorbance of 1.5. Similarly, Curve C in Figure 5 shows the result of using the factor (T>‘)'(‘“bl+“°3)/2 to correct the analytical curve of quinine sulfate for primary absorption interference. Again, the simple exponential cor- rection compares well with Equation (3.13) (Curve 8) at low absorb bance but gives a positive error above an absorbance of about 0.5. The reason for this behavior is easily seen from the series expansion of either correction factor. For example, at low absorbance, (OB-Ga)£nTA. ~ (oB-oa)2nTA. 9 (TA.)08 - (Tx') 0 (1+982nTX'+%O§£n2TX')'(]+®a£"TX'+%G§£n2TA') , (3.14) (GB'Oa)2nTAI 1 2 2 2 (3'15) (GB'ea)£nTX'+2(GB'Ga)£n TA. 1' 1 1 (3.16) -(0 +0 )/2 (7A.) 0' B (3.17) Figure 5. 50 Fluorescence analytical curve of quinine sulfate A. Blank and source corrected fluorescence, wB = 0.451, “b = 0.246. B. Primary absorption corrected fluorescence using the factor (we-wa)£nTA (1,)96 - (TA)wa C. Primary absorption corrected fluorescence using the factor (TA)'(wbfi”B)/2. 51 500 — 45 C) (D 1 300 '— FLUORESCENCE INTENSITY AT 450 NM C B 200 ‘- / A I l I L 05 ID 1.5 2.0 QUININE SULFATE CONCENTRATION (mm Figure 5 52 Thus, the simple exponential corrections for primary and secondary absorption are low absorbance limiting forms of the expressions which have been derived here and are in error when cubic and higher order terms of the series expansion become significant. One would expect the magnitude of this error to increase with the width of the window. This has been observed experimentally. CHAPTER IV THE EFFECT OF REFLECTIONS ON RIGHT-ANGLE MOLECULAR FLUORESCENCE MEASUREMENTS A. Introduction Although the treatment presented in the previous chapter assumes that reflections within the sample cell have a negligible effect on right-angle fluorescence measurements, some simple considerations suggest that this is usually not true. For a typical glass sample cell, four percent of the exciting radiation transmitted by the sample solution will be reflected back into the sample by the air-glass interface of the cell wall. The reflected light will excite the sample and contribute an additional component to the measured fluores- cence signal. Similarly, a small fraction of the fluorescence radia- tion initially emitted in a direction away from the fluorescence detector will be reflected back to the detector to be measured. Logically, the magnitude of these effects will depend on the amount of radiation transmitted to the reflecting surfaces and, therefore, on the sample absorbance. Because these reflection effects could influence the accuracy of the absorption correction procedure, a mathematical model of the reflection phenomena would be valuable so that corrections could be applied. In this chapter, expressions are developed to describe the 53 54 effects of the natural reflection properties of the sample cell in a dispersive right-angle fluorimeter on the measured fluorescence signal. The primary and secondary absorption correction factors described previously are modified to account for reflection effects, and an improvement in the accuracy of the absorption correction pro- cedure resulting from this development is demonstrated. Although equations equivalent to those given here have been proposed by van Slageren §t_gl, (50), this work is presented to clarify the theo- retical basis of the expressions and to offer experimental evidence for their validity. Reflections of the exciting radiation are con- sidered first. 8. Theorv 1. Reflections at the Exciting Wavelength To understand the manner in which reflections of the exciting radiation affect the measured fluorescence signal it is easiest to represent the fluorescence photoanodic current as the sum of an infinite number of current components which are generated by suc- cessive passes of the exciting radiation through the cell. Absorp- tion and reflection effects at the emission wavelength can initially be ignored because they are totally independent of the exciting wave- length. If all four cell walls are assumed to be identical, the fluorescence photoanodic current from an instrument of the type described in the previous chapter can be represented by 55 (1,)“8 - (1,)wa (T,)"wa - (T,)"ws 'fpa = (1'8r)[i znTA ) + PrTX ( inTx ) + w w l-w 1-w 2 2((TX) B - (TX) 0) + T 3p 3((T)\) a - (TX) 8) + ,,] .Jo T o X r znTA X r InTA (4.1) where all symbols not defined previously are explained in Table 2. The first tenn in Equation (4.1) represents the signal component generated on the initial pass of the excitation beam through the cell, as described in Chapter III. The second term represents the signal component generated by the first reflection of the excitation beam. Note that the window parameters for this term are different because the direction of excitation is reversed. The factor prTX accounts for absorption of the exciting beam on its initial pass through the cell and for the efficiency of the reflection. Success- sive reflections generate the other terms shown by similar reasoning. Equation (4.1) can be factored to give Equation (4.2). - - J°(]-pr) mg ”a + { I-wu (T )l-wB lfpa - W {(TA) " (TX) } DY'TX (TA) " A } X (1 + przTAZ + pr4TA4 + ...) (4.2) Then, by applying the identity 2 xn = l/(l-x) and further factoring, n=0 56 Table 2. Definition of Symbols in Equation (4.1). ra + rb ' 2rarb l - rarb reflectance of the cell wall at the excitation wavelength, dimen510nless reflectivity of air—cell interface at the excitation wavelength, dimensionless (na - nc)2 (na + n)2 refractive index of air at the excitation wavelength, dimensionless refractive index of cell wall at the excitation wavelength, dimensionless reflectivity of the cell-air interface at the excitation wave- length, dimensionless (nc - na)2 2 (nc + "a) 2.303AijgjgfixFexfofA.FemA.Sx.dXdX' (See Table l.) 57 Equation (4.2) can be reduced to Equation (4.3). U.) 8 w - - (TA) ’ (TA) a 1 + pr(TX)2 wB wa 2 2 lnTA 1 - pr TX (4.3) 1.fpa = (l'pr) The second factor in brackets in Equation (4.3) describes the effect of reflections at the excitation wavelength on the measured fluores- cence signal. This expression predicts that the magnitude of the reflection effect will depend on the sample absorbance, as reasoned above, on the efficiency of the reflection, and on the excitation window parameters. 2. Reflections at the Emission Wavelength By the same reasoning and approach used above, Equation (4.3) can be modified to account for the effects of reflections of the fluores- cence radiation. This is shown in Equation (4.4). (TA)wB - (Txlwa (1'0r)(' + pr(TA)2-wB-wa 1 = (1-p ') fpa _ 2 2 r RNTA 1 pr TX 0 G 1-9 _ 1-@ (TA') 8 " (TA') 0. + p '1' (TA.) a (TA') 8 (4.4) AORnTA. r A AORnTA. 2 ”was - (1111901 3 (Twp-e, - (T . + (D 'T I) T (D 'T I) ... r X Y‘ X 0 AOZnTA. AOEnTA. 58 The quantity pr' is the reflectance of the cell wall at the emission wavelength. The odd terms of this sum arise from fluorescence radia- tion initially emitted toward the fluorescence detector. The even terms are generated by fluorescence radiation initially emitted away from the detector. Obviously, the series in Equation (4.4) is of the same form as that generated by reflections of the exciting radiation. It follows that Equation (4.4) can be reduced to the form: 1.fpa ' 2- - ‘1 e prB- (will (1-o,)(1+o.(T,) “'8 ”a (1p) 8- (we 2 2 ._ WT) J 1 - or T, AOILnT)‘. (T-p,')(1 + pr'(TA-)2'QB'OG I 2 0 1 - (pr TA.) (4.5) Taking the limit of this expression as TX and TX. approach 1, one obtains pra = (we ‘ wa)Jo (4'6) which is consistent with the definition of the absorption-free fluores- cence signal defined in Chapter III. From Equations (4.5) and (4.6), a correction factor for absorption and reflection of exciting and fluorescence radiation can be defined as 59 2 2 1 .. AIDXJIT I Dr T)‘ AGQNTAI f ar (TA1we - (TA1wa (1-o,)(1+o,(1,12'88‘wa) (1,.198 - (T,.)Oa 1 ' (pr' TX')2 (4.7) (l-or')(1 + or'(Tx.)2'eB'ea) 3. Detennination of the Cell Reflectance To apply Equation (4.7) it is necessary to know the values of the cell reflectance at the excitation and emission wavelengths. In most experimental situations these parameters can be detennined from the transmittance of the cell filled with solvent measured with respect to air at the appropriate wavelength. If the cell walls do not them- selves have an appreciable absorbance, the intensity of radiation transmitted by the solvent filled cell is given by 2 I = IOITS(1-or)2 + T5311-o,)2or + T55(1-pr)20r4 + ...1 (4.8) = 1015(1-pr)2 [1 + (Tspr)2 + (Tspr)4 + ... ] (4.9) and 2 T 1- 1/10 = 5( or) 2 (4.10) 1 - (Tsar) where I0 is the incident radiation intensity, I is the transmitted radiation intensity, and TS is the transmittance of the solvent. In 60 most cases. TS will have a value close to unity so that Equation (4.10) reduces to r (4.11) and or is given approximately by _ l - (I/Io) - (4.12) l + (I/Io) or C. Experimental Fluorescence measurements used to test the validity of Equations (4.5) and (4.7) were obtained with a spectrofluorimeter specially con- structed to allow precise and accurate positioning of the sample cell with respect to the excitation and emission windows and to allow accurate, direct measurement of the excitation and emission window parameters. Details of the construction, calibration, and operation of this in- strument are given in Chapter V. Transmittance measurements for the absorption corrections and reflectance detenminations were obtained with this same instrument by positioning the reference detector to allow the excitation beam intensity transmitted by the sample cell to be measured. Appropriate bandpass filters were placed between the cell and the detector to reduce stray light interference. 61 0. Results and Discussion To verify Equations (4.5) and (4.7), solutions of 2 x 10'5 M quinine sulfate plus 0 to 5 x 10'5 M fluorescein in 0.1 N H2504 were prepared. The solutions were excited at 365 nm and the fluores- cence of quinine sulfate was measured at 436 nm. The transmittances of each solution and of the solvent were measured at the excitation and emission wavelengths by the procedure noted above. Correction factors were calculated from Equation (4.7) with reflectance values computed from Equation (4.12). Correction factors were also computed using reflectance values of zero to illustrate the effect of ignoring the reflection corrections. The average corrected fluorescence in- tensities by each of these procedures are listed in Table 3 along with the experimental standard deviations computed from three measure- ments. The data show that reflections do have a significant effect on the measured fluorescence intensity and on the accuracy of the absorption correction procedure. For the case in which reflection effects were ignored in the correction calculations, the corrected fluorescence intensity decreased to about 96 percent of its low ab- sorbance value when the secondary absorbance of the sample was 0.35 or greater. When the reflection corrections were included, however, the corrected fluorescence intensity remained constant to better than one percent up to an absorbance of 2.12, as it theoretically should. At low sample absorbances, Equation (4.5) predicts that multiple reflections within the sample cell will exactly compensate for the excitation radiation lost due to reflection as it enters the cell. It is also predicted that multiple reflections will compensate for 62 Table 3. 'Test of Reflection Corrections. Corrected Fluorescence 365 A436 0r = Pr' = 0 or = 0.0407, pr' = 0.0411 0.126 0.000 2856 i 48 2869 i 48 0.145 0.353 2765 s 11 2867 s 11 0.163 0.701 2750 s 25 2877 s 27 0.183 1.060 2732 i 22 2866 s 23 0.202 1.408 2739 s 10 2877 s 11 0.222 1.767 2708 i 43 2847 s 45 0.241 2.122 2740 i 29 2882 i 31 63 the fluorescence radiation lost due to reflection as the emission beam leaves the cell. At higher absorbances, however, the signal components generated by multiple reflections are attenuated by the absorption processes causing the measured fluorescence signal to be too small. The error should be of the same magnitude as the cell reflectance, j;g;, if or' = 0.04, the error should be four percent. The data in Table 3 confirm this prediction. It is clear from these results that by including corrections for multiple reflections at the excitation and emission wavelengths the accuracy of the absorption correction procedure is significantly improved. The results presented here are more accurate than those given in Chapter III because reflection effects have been explicitly corrected, while in Chapter III they were removed by a curve fitting procedure. Although the errors caused by reflection phenomena are negligible at low sample absorbances, they could be as high as eight percent for a typical sample cell if the sample absorbances at the excitation and emission wavelengths are both 0.35 or greater. Because of this, reflection effects should be taken into account in absorption- corrected right-angle fluorimetry. CHAPTER V AN AUTOMATED INSTRUMENT FOR ABSORPTION-CORRECTED FLUORESCENCE MEASUREMENTS BY THE CELL SHIFT METHOD A. Introduction In Chapters III and IV it was shown that the nonlinearity of the signal-concentration relationship in right-angle fluorimetry, due to reflections and absorption of exciting and fluorescence radiation by the sample, can be corrected with a simple mathematical treatment. The measured fluorescence intensity of the sample is multiplied by a correction factor to produce a corrected intensity that is a linear function of the fluorophore concentration even when the sample ab- sorbance is as high as 2.0. The correction factor depends primarily on geometric instrumental parameters, the sample transmittances at the excitation and emission wavelengths, and reflectance parameters of the cell. Transmittance values needed to compute the correction factor can be obtained with a standard laboratory spectrophotometer. How- ever, this procedure suffers from the following problems: many fluorescent samples undergo spectral changes during the delay between the fluorescence and transmittance measurements; analysis time is increased due to the need for additional measurements; and care must be taken that the spectral bandwidths of excitation and emission are 64 65 closely duplicated by the spectrophotometer (69). Ideally, to avoid these problems, the sample transmittance should be measured simul- taneously with its fluorescence and with the same instrument so that the same optical system is used. Simultaneous measurement of sample fluorescence and transmittance at the excitation wavelength has been accomplished in a right-angle fluorimeter by using an oscillating mirror assembly to move the beam of exciting radiation continuously between sample and reference cells and to direct the transmitted radiation to an optical detector (71). The sample transmittance at the emission wavelength must still be determined separately, however. Further development of this design to allow the transmittance at the emission wavelength to be determined has not been attempted because of the complexity of the optical system that would be required. A unique approach has been recently proposed to overcome this problem. It has been shown (72) that in a right-angle fluorimeter the sample absorbance at the excitation wavelength can be derived from fluorescence intensities measured at two locations along the path of the excitation beam as it passes through the sample cell. The cell is manually moved in a direction parallel to the excitation beam, and fluorescence intensities are obtained at two cell positions. Similarly, the sample absorbance at the emission wavelength can be determined by shifting the cell in a direction perpendicular to the excitation beam to cause a change in the pathlength of the fluores- cence radiation through the absorbing solution. The technique is known as the method of "cell shift" (69). 66 The cell shift approach appears to be an effective and easily implemented method for the correction of absorption errors in right- angle fluorimetry. However, if the sample cell is positioned man- ually, the procedure can be tedious, time consuming, and subject to significant errors. To reduce these problems and to facilitate further evaluation of the cell shift technique, an absorption—cor- rected spectrofluorimeter was developed in which a dedicated micro- computer is used to position the sample cell automatically for the cell shift measurements. The microcomputer also integrates the fluorescence and source (reference) signals, scans the excitation and emission monochromators, and transmits the acquired data to a minicomputer for data reduction and permanent storage. In this chapter the basic design and operation of this instrument are des- cribed. The theories developed in Chapters III and IV are used to explain the principle of the instrument operation. It is shown that corrections for multiple light reflections are necessary to determine the sample transmittance accurately by the cell shift technique. And finally, the utility of the instrument is demonstrated in an application to the fluorimetric determination of aluminum in the presence of iron. B. Theory Figure 6 illustrates the concept of the cell shift method. Baffles are located in the fluorimeter so that only a small, fixed volume of space is excited by the radiation source and viewed by the fluorescence detector. The sample cell is moved to bring the 67 ORIGIN POSITION 1 13 EM E] EM l (____, .— Y ex ex POSITION 2 POSITION 3 E]; E. E . E)( E)( Figure 6. Cell positions for the cell shift method. EX = excitation window, EN = emission window. 68 sample solution into this region (position 1) and to vary the thick- ness of solution through which the exciting radiation must pass (posi- tion 2). Similarly, the cell is moved to vary the sample thickness through which the fluorescence radiation must pass to reach the detector (position 3). In accord with the work described in Chapter IV, the fluores- cence intensities measured at cell positions 1, 2, and 3 can be described by the expression: 00 l e e Z-w -00 _ (”x1 B"-(TA)Dan][(TA.) Bil-(TX) an][(1-pr)(1+pr(‘r>‘) Bn 611)] F - F " ° L((0 )tnTA (eBn-oanmlx. 1'pr2 T12 J Bn'won 2-0 -0 [(l-pru )(T+o,.. (1,.) B" “"1] 1'(er TA')2 where n designates the cell position (n = l, 2, or 3), F0 is the ab- sorption-free fluorescence intensity of the sample, and an’ nan, 980’ and 0b" are the instrumental window parameters for the cell position of concern. As for Equation (3.10), Equation (5.1) is valid if the excitation and fluorescence beams are collimated and of narrow spectral bandwidth compared to the sample absorption bands. Stray light, scattering phenomena, and reemission of absorbed fluores- cence must also be negligible. When the fluorescence cell is shifted from position 1 to position 2, the excitation window parameters are not changed, i.e., 0 = 0 81 82 69 and 0a] = 902' However, the emission window parameters are each de- creased by an amount on so that w02 = ma] - 5m and ”82 = ”81 - Ow. Similarly, when the cell is shifted from position 2 to position 3 the emission window parameters are not affected, ”83 = ”82 and w03 = ”82’ but the excitation window parameters are increased by an amount 60 so that 983 = 082 + 60 and 0G3 = 902 + 60. From these rela- tionships and Equation (5.1), it can be shown that TX is related to the measured fluorescence intensities F1 and F2 by the nonexplicit relationship: Z-w -w 1 + Or(TA) 82 a2 I/Ow TA = (Fl/F2) (5.2) Z-w -w 1 + pr(TX) B] a] If reflection effects are small, however, TX is given approximately by TX 2 (Fl/F2)1/Ow (5.3) By substituting this relationship into Equation (5.2) to give Equa- tion (5.4), a more accurate estimate Of TX is obtained. (Z-w -w )/Ow 1+0r(F1/F2) 82 02 1/6w TA = (Fl/F2) (2-081-uh11/80 (5.4) 1+pr(F1/F2) By a similar argument, it can be shown that TX' is given by 70 (2-0 -0 )/oo l/oo 82 02 1+0 I F /F TX' = (Fs/Fz) (2-083-Oa3)/60 1+Drt(F3/F2) The instrument described below automatically measures F1, F2, and F3 by shifting the fluorescence sample cell as shown in Figure 6. From known values Of or and or. and the window parameters for each cell position, TA and TA. are calculated from Equations (5.4) and (5.5), and Equation (5.1) is solved for the value of F0. C. Instrument Design 1. General Description The major components of the absorption-corrected spectrofluorim- eter and their physical arrangement are shown in Figure 7. Because of the spectral bandwidth and stray light limitations on Equation (5.1), monochromators are used for both excitation and emission wave- length isolation. GCA McPherson model EU-7OO monochromators are used at their maximum slit widths (2 mm) and maximum spectral band- widths (4 nm). A 150 W xenon or a 200 W xenon-mercury arc lamp is used as the excitation source. In the cell compartment, radiation from the excitation monochromator is mechanically chopped at 500 Hz and collimated by a single convex quartz lens (1.5 inch dia., f/3). A quartz beam splitter directs a fraction of the collimated beam to a source reference detector, a rhodamine B quantum counter and RCA 1P28 photomultiplier tube. The remainder of the excitation beam is directed to the cell (type A-23 UV quartz, Markson Science, Del Mar, CA) 71 .pcosmmcmccm grasp new mpcwcogEoo pcmsacpmcH .n mczawm ...!a cards—$100202 ...zwstkdaznuo ghdiomzoozoz momaom owJOOo zo.wm_—zm 1.1. mo 2054.588 mxé: cwhhjmm Zdwo // fl \\.. cmaaoro \ jwu o /. «HP—.2300 23.5430 72 and a slit arrangement like that shown in Figure 6. The excitation and emission slits are each 5.0 mm tall and 2.4 and 2.0 mm wide, respectively. The rather small dimensions chosen for the slits are necessary to maintain a high degree of beam collimation, even though this severely limits the sensitivity of the instrument. The emis- sion slit and the entrance slit of the emission monochromator define the emission beam. These two slits are separated by a distance of 11.0 cm. With this arrangement, a lens is not needed to collimate the emission beam. The fluorescence detector is a Hamamatsu R666 photomultiplier that is bathed in cold N2 (74) to reduce the dark cur- rent shot noise. Movement of the fluorescence cell to the positions shown in Figure 6 is controlled by a microcomputer with the aid of an x-y positioner constructed in our laboratories. For all work described in this chapter, the displacement between cell positions 1 and 2 and positions 2 and 3 was 0.650 cm. The window parameters for each cell position were those given later in Table 4. 2. The Microcomputer The computer interface of the instrument is shown schematically in Figure 8. Data acquisition, cell position, and the monochromator wavelength drives are under direct program control of an Intersil IM6100 12-bit microprocessor. The IM6100 CPU, 8K Of RAM, memory extender, and two serial ports were purchased in kit form from Pacific Cyber/Metrix (San Ramon, CA). One serial port is used to communicate with the serial port of a PDP 8/e minicomputer and the .xcozpm: empaasou Co Emcumwn xoopm .m mczawg 73 wuhmusw=g.mxxzuommzxyaau ClCK? I 1 Eat) 1v5 IUfiQOflMBUVCET 1:0 CL5F _ . _ T . _ 2 1 0 “Xv. x.dmfixqfig.vnkpo Axum P u 0. 0 Figure 22 108 measured fluorescence response due to absorption of the exciting light is evident. Figure 21A illustrates the accuracy of the cor- rection for absorption of exciting radiation by sample components other than the fluorophore. Corrected and uncorrected fluorescence intensities for a series of solutions of 2 x lO'5 M quinine sulfate and 0 to 4 x lo”4 M gentisic acid in 0.1 N H2504 are plotted against the solution absorbance at the exciting wavelength. In Figure 22A, raw and corrected fluorescence intensities for solutions of 2 x 10'5 M quinine sulfate and 0 to 6.5 x lO'5 M fluorescein in O.l N H2504 are plotted as a function of the solution absorbance at the emission wavelength. The uncorrected data in Figure 22A reflect the effects of absorption of exciting radiation by both the fluorophore and the sample matrix and the effect of absorption of the fluorescence emis- sion by the sample matrix. The straight line in part A of each figure represents the theoretical absorption-free fluorescence response which was determined by extrapolating from the corrected data at low ab- sorbance (50.0l). Part B of each figure shows the precision of the corrected results, as the relative standard deviation (RSD), as a function of the sample absorbance. In each of these tests, the cell shift corrected results do not differ significantly (a = 0.05) from the theoretical lines up to an absorbance of 2.0. In Figure 20A, the corrected results are in error by less than two percent up to an absorbance of 2.7. The decrease in accuracy at higher absorbances is most likely due to inaccuracies in the extrapolation of the theoretical fluorescence resnonse and in the values of the instrument window parameters used in the correction 109 calculations. No fundamental limitation of the accuracy of the cell shift method is indicated under the conditions of these experiments. As shown, the relative standard deviation of the corrected fluores- cence results was very good in each test, ranging from 0.2 to about 2.0 percent. Increases in the RSD occurred as the magnitude of the measured fluorescence intensities and the fluorescence signal-to- noise ratio decreased. This behavior is discussed in more detail later. 2. Effect of Reemission To investigate the accuracy of the cell shift method under condi- tions in which the reemission of absorbed fluorescence is not negli- gible, solutions of 10'7 to 5 x TO"6 M fluorescein in O.l N NaOH were prepared and excited at 3l3 nm. Fluorescein has a relatively small Stokes shift of llOO cm"1 (79) and, therefore, a large degree of overlap between its absorption and fluorescence emission bands. Fluorescence emission was monitored at 5l0 nm (in the overlap region) and also at 540 nm (outside the overlap region). In Figures 23 and 24, the corrected fluorescence intensities at each emission wavelength are plotted against the fluorescein concentration. Also plotted are values of the apparent solution absorbance calculated by the cell shift method and used to compute the absorption correction factors. The straight lines in each figure are absorbance analytical curves determined from spectrophotometric data and extrapolated fluorescence analytical curves. The data show that at concentrations greater than about l0"5 M there is a definite decrease in the accuracy of the 110 400 ::4:F::}4;:}%::}:;% 3.0 350 -*- E «F 1-2.5 fi 300 -— ‘1 :2 " --2.0 E 250 -— . E 200 —— . --15 m -. 0 150 —- . + é _. - -—1.0 g 100 -- ‘ 1- : -- A ‘1“‘(L5 ‘3 50—- -.- . qb _.' l l I l I l J l l 1 I l l 1 I l I l O 'I I I I I 1 1 1 r I I I I I l j T 1 000 0.0 1.0 2.0 3.0 4.0 5.0 FLUORESCEIN CONCENTRATION, M x 104 Figure 23. Effect of reemission on the corrected fluorescence (A) and apparent sample absorbance at the excitation wave- length (I) when fluorescence is monitored outside the overlap region. "N Elf 1V BMWSQV FLUORESCENCE INIENsmr AT 510 NM 3 0 Figure 24. 111 l l 1 I J l 1 I l 1 1 I 1 1 1 I 1 1 l J I I I I I I I I I I I ‘l I I I I r I I . 3 O "’ O .- , 2.5 " ~-2.0 .. O I ah- .. O I '1" I O —— . .4. 4L " ‘ +0.5 «4- 4F '~ 1 l I l 1 1 I l l 1 I 1 1 1 I 1 1 1 1 I 1 l I fl I ' I I T 1 T I I I r I I 000 0.0 1.0 2.0 3.0 4.0 5.0 FLUORESCEIN CONCENTRATION, M x 104 Effect of reemission on the corrected fluorescence (A) and apparent sample absorbance at the excitation (I) and emission (O) wavelengths when fluorescence is moni- tored in the overlap region. somaosav 112 absorption-corrected fluorescence. Two effects are apparent from Figures 23 and 24. First, the presence of a small reemission component in the measured fluorescence signals causes a negative error in the sample absorbance calculated by the cell shift method. This is similar to the effect of stray light on a spectrophotometric absorbance measure- ment. The absorbance error tends to cause a negative error in the cor- rected fluorescence. Opposing this effect, however, is the reemis— sion component in the measured signal which tends to make the cor- rected fluorescence larger than the theoretical value. The nature of the net error due to reemission depends on the emission wavelength chosen. Away from the region of overlap (Figure 23), where correc- tions for absorption of fluorescence radiation are minor, the absorp- tion-corrected fluorescence shows a positive error. The error in- creases in magnitude as the fluorophore concentration is increased. If fluorescence is monitored in the region of overlap (Figure 24) where the correction for absorption of fluorescence radiation can be significant, however, the error in the corrected fluorescence is first positive, but eventually decreases and becomes negative as the fluorophore concentration is increased. Evidently, the combined negative errors in the apparent sample absorbances at the excitation and emission wavelengths become large enough to dominate the effect of the extra component of fluorescence due to reemission. Further evidence of this is given in Figure 25. Absorption-corrected fluores- cence emission spectra for 5 x 10's, 10's, and 3 x l0'5 M solutions of rhodamine B in ethanol, normalized for concentration, are shown. Clearly, as the fluorophore concentration is increased, a decrease in 113 E 0.030» .202?- 2 9 a: I a ‘ a E on- :g o. I o A a In 00° 0. o 2 0.020-'- ‘n 0. II o x. ‘ - 00 o g -.. 2 O. a d O U 0‘ a a. 0:0.0 g 0.0IO-"' 2° 9:... 03 a. g 2' .g I... u ‘ O '9 g . ...:I I g "" :0 33.: a" g. 0.000 530 550 570 590 610 630 WAVELENGTH, NM Figure 25. Absorption-corrected fluorescence emission spectra of rhodamine B in ethanol. (0) 5 x l0"6 M, (A) 10'5 M, (D) 3 x 10'5 M. 114 fluorescence intensity on the low wavelength side of the emission band and an increase in intensity on the high wavelength side occur which are statistically significant. This is exactly the effect predicted from the data of Figures 23 and 24. Regardless of the emission wavelength chosen, the accuracy of the cell shift method can be poor when the absorption and emission bands of the fluorophore overlap. The data of Figures 23 and 24 and of Figure l9 in Chapter V indicate, however, that errors due to reemission are insignificant below a concentration of about l0'5 M. Because the examples used here approach the worst case for degree of absorption-emission overlap, it is expected that, in general, re- emission errors will not be significant below this concentration limit and below a much higher concentration limit for many fluoro— phores. 3. Effect of Light Scattering Scattering of the exciting and fluorescence radiations by the sample matrix is a potential problem common to all fluorimetric methods. To determine the effect of matrix scattering on cell shift corrected fluorescence measurements, a series of solutions of 4 x 10'5 M quinine sulfate in O.l N H2504 containing increasing amounts of soluble starch were prepared. The solutions were excited at 365 nm and fluorescence was monitored at 450 nm. In Figure 26 the corrected fluorescence intensities and solution absorbances, normalized to the values in the absence of the scattering agent, are plotted as a function of the scattering agent concentration. Clearly, the effect 115 IHZO dr- unl— ‘ cull— .- —— ‘- ‘ Oi 11M? AflflRWUUJZED’lflfiRHEIHWfl) FLLKNQESKXHVCET .JCMflflflHOEEflV.&HVlde'CEHHWVMRKNV 0.80 : 1 : I : 1 1 0.8 0 1 2 3 4 SCATTERING AGENT CONCENTRATION, REL. UNITS Figure 26. Corrected fluorescence (2:) and apparent absorbance (E3), normalized to zero scattering agent, of a constant con- centration of qUinine sulfate as a function of increasing amounts of soluble starch. 116 of light scattering by the sample matrix is a positive error in both the apparent sample absorbance and the corrected fluorescence. The magnitude of the error increases as the scattering agent concentration is increased. 4. Effect of §pectral Bandwidth Because the spectral bandwidths of excitation and emission have been shown to affect the accuracy of other absorption correction pro- cedures (66), the effect of this measurement parameter on cell shift corrected fluorescence measurements was investigated. The excitation monochromator of the spectrofluorimeter, which has a maximum band- width of 4 nm, was replaced with a Corning 7-60 band filter which has a bandwidth at half height of approximately 40 nm. Quinine sul- fate standards from l0"7 to 4 x lO'4 M were excited with a 200 w Xe-Hg arc lamp, and their fluorescence emission was measured at 450 nm. In Figure 27 the corrected fluorescence intensities and solution absorbances determined by the cell shift method are plotted against the quinine sulfate concentration. The extrapolated fluorescence response from low concentration and the absorbance analytical curve determined by spectrophotometry are also shown. Clearly, the increased excitation bandwidth resulted in negative deviations of both the sample absorbance and the corrected fluorescence from their theo- retical values at concentrations above l.5 x lO'4 M. The effect is similar to that which polychromatic radiation has on spectrophoto- metric measurements. Although the effect on the corrected fluores- cence in this example is not very severe, it would be expected to 117 400 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3.0 350 ~— § .. 2.5 —n— .1.- § 300 . a 2; " . -h-1LC) g3 E 250 -- U) "" o 1 g —— ...1. . M E 200 .. . . 15 2‘ MJ ‘ " c4 0 150 -— g __ --1.0 2 g 100 -- 1- % " a—a.5 50-— d.- 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0.0 0.0 1.0 2.0 3.0 4.0 OUININE SULFATE CONCENTRATION, M x 1 04 Figure 27. Corrected fluorescence (25) and apparent absorbance at the exciting wavelength ([3) vs. concentration for quinine sulfate excited with a Xe-Hg arc lamp and Corning 7-60 filter. 118 be worse if a continuum radiation source had been used or if the emis- sion monochromator had also been replaced with a band filter. By using interference filters rather than band filters the errors would probably be less severe. 5. Correction Precision Although Novak (69) has discussed the effect of the instrument window parameters on the precision of the absorption correction factors, the results shown in Figures 203, 2lB, and 228 are better understood through a simple extension of his discussion. Ignoring the corrections for reflections within the sample cell, the corrected fluorescence intensity from the cell shift method Fc is given, to a good approximation, by FC = 1'1" Fzy F32 (6.l) where F1, F2, and F3 are the measured fluorescence intensities at the three cell positions, and the exponents x, y, and z are given by X = -(w82 + wa2)/2 annn ‘° 11? A! AU AAAAAA ...... no»)... vvvvvv a X ” aunt-tn“ ”...-U IOUI’U‘ vvvvvv O X 0 DDDDDDDDDDDO ”uHHMHHuH-IHI‘ 90009008.”... (89) XTC 0X10 «as» EFIEI "‘ Figure Al. Data Bus Buffering and Address Circuitry x5325 mficouwo EB foam: venom 5E .~< ocsaca ..HUFIO—u 933399.99 DOOBODDDD vb-l .- Ii-C'ICFIO—a .4. ..u- ¢<¢¢ 139 .>pr:ocwu cowpmpzcmm escape: use pcsccmucH chma Focpcou .me mtsmcu Luau... 1U. h20¢xa hIthu :\¢ mum. hunuc n.+ “a“.o.. an- “mason. mun .ov anhv atkv Aodv auflv asfiv nouv 140 Table A1. Identity of Circuit Components in Figures Al-A4. Circuit Symbol Device Identity U1, U2 U3, U4, U5 U6, U7 U8, U9, U10 U11, U12, U13 U14, U15 U16, U17 U18 U19 U20 U21 U22 U23 74LS366 74175 74L5138 2708 7489 74125 74365 7404 7408 7402 7403 74121 74123 Tri-state Hex Buffers Tri-state Quad 0 Flip-Flops Decoder Erasable Progranmabl e Read-Only-Memory 64-Bit Read/Write Memory Tri-state Quad Buffer Tri-state Hex Buffer Hex Inverter Quad 2-Input AND Gate Quad 2-Input NOR Gate Quad 2-Input NAND Gate Monostable Multivibrator Dual Monostable Multivibrator 141 nt Lavout Figure A4. PROM Board Comoone 142 enabled by address line A0, and therefore by any address in the octal memory space 4000-7777. Only four bits of its memory contents are accessed at one time, however, because of the action of tri-state buffers U14 and U15. When addresses of the form 4XXX and 5XXX are decoded, EPROMs U9 and U10 and buffer U15 are enabled. When ad- dresses of the form 6XXX and 7XXX are decoded, EPROMs U8 and U9 and buffer U14 are enabled. In this manner, 2048 12-bit words of EPROM are obtained from only three memory devices. The PROM monitor board also contains 16 words of RAM for scratch pad operations by the monitor. This memory is arranged in three 16 x 4-bit arrays, devices Ull-U13. Device U11 contains the most significant four bits of memory. Device U13 contains the least significant f0ur bits. Devices U6, U20, and U7 enable the RAM for read and write operations when addresses 0000-0017 are decoded. 2. Read and Write Operation The manner in which the IM6100 CPU reads from and writes into the memory on the PROM board can be understood by referring to Fig- ures A1 and A2. When power to the microcomputer is on, tri-state buffers U1 and U2 continuously drive the contents of data bus lines DXO-DXll onto the PROM board where they become data input lines DIO- DIll.. Early in each instruction cycle, a 12-bit memory address is placed on these lines by the IM6100 CPU. At the same time, a pulse is generated on the LXMAR control line. The trailing edge of this pulse latches the address into the D flip-flops U3-U5. If the PROM board memory is to be accessed, a short time later the CPSEL 143 line is pulsed L0 by the CPU. This causes devices U6, U7, U19, and U20 to decode the address and to enable the appropriate memory devices so that their contents appear on the data output lines DOO-DOll. During the first half of the instruction cycle the control line XTC is exerted HI. This prevents data from being written into the RAM devices U11-U13 and in combination with CPSEL being LO en- ables tri-state buffers U16 and U17 to drive the memory contents onto the DX lines where they are read by the CPU. If a write operation into the scratch pad RAM is to be carried out, the read sequence described above is first executed but is irrelevant. After the read operation, the CPU exerts the XTC line LO which write enables the RAM devices Ull-U13. A short time later, the data to be stored in memory are placed on the DX lines and CPSEL is exerted LO. If the appropriate memory space is decoded, the chip enable pins of Ull-U13 are exerted L0 by U7. The combination of chip enable and write enable signals causes the contents of the 01 lines to be transferred into the RAMs. 3. Control Panel Interrgpt Circuitry Memory on the PROM monitor board is selected by the IM6100 CPU over the microcomputer mainframe memory when the CPU exerts the CPSEL line LO rather than the MEMSEL control line. To enter the PROM monitor the CPSEL line must be activated by issuing a control panel interrupt request to the CPU. This action is carried out by the circuitry shown in Figure A3. When power to the microcomputer is turned on, monostable U23 automatically pulses the RESET control 144 line LO for two seconds. This forces the CPU into the HALT state, clears the accumulator, and sets the program counter to 7777. The trailing edge of this pulse triggers another two second logic LO pulse to the CPREQ line to request the control panel interrupt. The second logic pulse also triggers device U22 to apply a logic LO pulse to the Run/Halt control line. This forces the CPU into the RUN state beginning at memory location 7777. If the IM6100 CPU is in the RUN state and is executing instruc- tions from mainframe memory, the PROM monitor can be entered by initiating the sequence described above manually or from program control. The sequence can be initiated manually by pressing the momentary action switches "RESET" and then “KBE” on the PROM board. If pin 1 of device U23 is connected to bus line 77 by a jumper, the PROM monitor entry sequence can also be started by executing a device input/output instruction that is decoded onto line 77 by another interface. A 6525 instruction was used for this purpose in this work. If a device interrupt or direct memory access request is being serviced by the CPU, both of these modes of PROM monitor entry are inhibited. C. Software Description The expanded IM6100 PROM monitor software closely resembles the original monitor program deve10ped by Joseph (75). The reader is referred to the original work for a detailed description of the program Operation. Features which have been added to the monitor include: 1) the ability to set the microcomputer instruction and 145 data fields by typing the keyboard command ”FN” where N is the octal field number; 2) the ability to open and examine successive and preceeding locations of mainframe memory with the keyboard commands "<" and ">"; and 3) the ability to return to the transparent mode from the local mode by typing the keyboard command "control-T". A further description of these commands and how they were incorporated into the monitor program is given in the program assembly listing on the following pages. I The monitor software was written in PAL8 assembly language and stored in the file MON.PA. After compiling the program to binary format, it was written into the EPROM devices with the aid of the C. G. Enke 2708 PROM programmer and the RSX-ll version of the program PROM. Since the expanded monitor software required less than 1K of EPROM memory, it was necessary to program only two EPROM chips. Because the program entry point is at location 7777, the program was written into devices U9 and U10 at chip addresses 1000-1777. 07000 07001 07002 07003 07004 07005 07006 07007 07010 07011 07012 07013 07014 07015 07016 07017 07020 07021 07022 07023 07024 07025 07026 07027 07030 07031 07032 07033 07034 07035 07176 7000 6136 6131 6146 6141 0000 0002 0004 0005 0006 0007 0010 0011 0012 0013 0014 0015 0016 0017 7000 7300 1235 3003 6031 5216 6036 1377 7450 5633 1376 6146 6141 5213 5200 6131 5203 6136 1231 7450 5634 1232 6046 6041 5226 5200 7600 0200 7200 7600 5401 0201 146 /////////////////////////////////////////////////// \‘\\‘\\ ////////////////////////////////////////////////// ORG= PDPIN= PDPISF= PDPOUT= PDPOSF= TEMP= FLDSW= PC= FLDFLG= SUBR= PCSAVE= CHKSM= WORD1= WORD2= BEGSW= RUBCT= BYTSV= FLG= CHAR= PROM MONITOR FOR IM6100 MICROPROCESSORS FILE: MON.PA 7000 6136 6131 6146 6141 'QOCIQIOQ 10 ll 12 I3 l4 l5 l6 l7 / /TRANSPARENT MODE / /SPECIAL CHARACTERS: \‘\\‘\\ MODE]. PDP , M200, K200, XMON. XMODEZ, K5401. 'TTY: CONTROL-A MAIN COMPUTER: 200(8) xORG CLA CLL TAD K5401 DCA FLDSW+1 KSF JMP PDP KRB TAD SNA JMP I XMDN TAD (201 PDPOUT PDPOSF JMP .-1 JMP MODE! PDPISF JMP TTY PDPIN TAD N200 SNA JMP I xnpppz TAD K200 TLS TSF JMP .-l JMP MODE! -200 200 MONITR MDDE2 5401 (-201 / / / / / / -> CHAIN 'I‘O nommn -) CHAIN TO BINARY LOADER /START OF MODE ONE /SET UP FIELD CHARGE SUBROUTINE ITTY HIT? /NO; CHECK PDP /YES; READ IN ICHECK FOR CONTROL-A AT TI'Y ICONTROL-A ? /YES; 00 TO MONITOR /NO; RESTORE CHARACTER /SEND TO PDP IWAIT PDPOUT FLAG /LOOK FOR MORE /PDP CHAR? /N0: CHECK ‘I'I'Y /YES: READ IN /SUBTR 200(8) /LEADER-TRAILER? /YES: 60 T0 BIN LOADER /NO. RESTORE xTD TTY /LOOK FOR MORE 07177 07600 07601 07602 07603 07604 07605 07606 07607 07610 07611 07612 07613 07614 07615 07616 07617 07620 07621 07622 07623 07624 INSTRUCTION 07625 07626 07627 07630 07631 07632 07633 07634 POINT 07635 07636 07637 07640 07641 07642 07643 07644 07645 07646 07647 07650 07651 07652 07653 07654 07655 07656 07657 07660 07661 07662 07663 07664 07665 07666 07667 07670 07671 07672 7577 7600 7300 3005 6132 6214 1321 3002 7040 3013 7040 3015 6131 5212 6136 3000 1000 2015 5252 0324 1300 7510 5242 7750 5237 1000 0323 1321 3002 1005 7650 4001 5210 2013 5311 5206 7200 2013 5266 3010 1000 3011 3015 5212 3012 1011 7006 7006 7006 7430 5263 7240 3005 7300 3013 5210 1011 7106 7006 7006 1012 / /BINARY /CHAINS / MODE2, START. MSBYTE, LSBYTE. FLDSET. LT. DAORG, GO. LSSAVE. OVER, DEPOST. 147 87600 LOADER SECTION TO MONITOR IF CHECKSUM ' 0 CLA CLL DCA FLDFLG 6132 RDF TAD DCA CMA DCA CMA DCA BYTSW PDPISF JMP LSBYTE PDPIN DCA TEMP TAD TEMP ISZ BYTSW JNP LSSAVE AND M0300 TAD K7600 SPA JNP K5201 FLDSW BEGSW DAORG SNA CLA LT TEMP M0070 K6201 FLDSW FLDFLG CLA SPA JMP TAD AND TAD DCA TAD SNA JMS JMP ISZ JMP JMP CLA ISZ JMP DCA TAD DCA DCA JMP DCA TAD RTL RTL RTL SZL JMP STA DCA CLA DCA JMP TAD CLL RTL RTL TAD FLDSW-l MSBYTE BEGSW END MSBYTE-2 BEGSW EPOST CHKSM TEMP WORDI BYTSW LSBYTE WORDZ WORDI OVER FLDFLG BEGSW MSBYTE WORDI RTL WORD2 IFALL THRU TO BIN LOADER IENABLE INITIAL FIELD CHANGE /INITIALIZE READER IGET CURRENT DATA FIELD /SET UP FIELD CHANGE SUBROUTINE IFORM -1 /SET BEGINNING-END FLAG ISET BYTE FLAG /WAIT FOR PDP FLAG /GET A CHARACTER /SAVE IT TEMPORARILY /SUBTRACT 200 /2 OR 300 ? INO. MUST BE AN ADDRESS OR AN /LT ? /YES /NO. MUST BE A FIELD SETTING /HOLD UP ON FIELD CHANGE AT THIS /ADD TO CHECKSUM REGISTER /SAVE IT AS WORDI /SAVE IT AS WORDZ 07673 07674 07675 07676 07677 07700 07701 07702 07703 07704 07705 07706 07707 07710 07711 07712 07713 07714 07715 07716 07717 07720 07721 07722 07723 07724 MODE 7430 5302 3404 4001 2004 7600 5305 3004 1004 3007 1011 1012 1010 5245 1011 7002 1012 7041 1010 7450 5722 7402 6201 7200 0070 0300 I48 SZL JMP ORIGIN DCA I PC JMS FLDSW‘I 182 PC 7600 JMP CHEX DCA PC TAD PC DCA TAD TAD TAD JMP GO TAD BS W TAD CIA TAD SNA JMP HLT 6201 MONITR M0070. 0070 M0300, 309 /KEYBOARD MONITOR SECTION / /TRIS SECTION OF THE PROM ROUTINE IS CHAINED TO /AFTER THE USER TYPES CONTROL-A WHILE IN THE TRANSPARENT /OR WHEN A ZERO CHECKSUM IS ENCOUNTERED AFTER.DOWNLOADING /AN ABSOLUTE BINARY PROGRAM FROM THE MAIN COMPUTER. / /KEYBOARD MONITOR TYPES A ”0” WHEN READY FOR INPUT / /INPUT IS OF THE FORM / IEXECUTE FIELD CHANGE /CLA /SAVE THE ORIGIN /SAVE THE NEW ORIGIN K7600, ORIGIN. CHEX, END. /CHECKSUM ZERO? /YES; CHAIN TO MONITOR K7402. /NO: ERROR HALT K6201. IQKL / NNNNX / /WHERE NNNN IS A FOUR DIGIT OCTAL NUMBER /AND X IS ONE OF THE FOLLOWING CONTROL CHARACTERS / / X = '/“ MEANS OPEN LOCATION NNNN AND / DISPLAY ITS CONTENTS / / X = 'D" MEANS DEPOSIT THE DATA NNNN / IN THE LOCATION WHICH IS CURRENTLY / OPEN / / X = “G" MEANS BEGIN EXECUTION AT THE / ADDRESS NNNN / /ONCE A LOCATION HAS BEEN OPENED IN THIS MANNER /THE FOLLOWING COMMANDS MAY BE USED '/“ = DISPLAY CONTENTS OF LOCATION WHICH IS CURRENTLY OPEN OPEN PRECEDING LOCATION AND DISPLAY ITS CONTENTS OPEN SUCCEDING LOCATION AND DISPLAY ITS CONTENTS “CONTROL-T" = RETURN TO TRANSPARENT MODE I< II II) II \\\\\\\\\\\\ 07200 07201 07202 07203 07204 07205 07206 07207 07210 07211 07212 07213 07214 07215 07216 07217 07220 07221 07222 07223 07224 07225 07226 07227 0723 07231 07232 07233 07234 07235 07236 07237 07240 07241 07242 07243 07244 07245 07246 07247 07250 07251 07252 07253 07254 07255 07256 07257 07260 07261 07262 07263 07264 07265 07266 7200 7300 1326 3006 1330 5750 1327 3006 1331 5750 33 3006 Q00 UV— 5750 3016 1325 3006 5746 7012 7012 3000 1331 3016 1333 3006 5746 7002 1000 3000 1334 3006 5746 7006 7004 1000 3000 1335 3006 5746 1000 3000 6031 5250 6036 6046 6041 5254 1337 7450 5751 1340 7450 5752 1342 7650 5274 /IN ADDITION. 149 /SET BY TYPING / / / /WHERE N / FN THE INSTRUCTION AND DATA FIELDS MAY BE IS THE OCTAL FIELD NUMBER. /IF ANY COMMAND SEQUENCE OTHER THAN THOSE LISTED ABOVE /IS TYPED. THE MONITER RESPONDS WITH A “7' AND IGNORES /THE COMMAND. / MONITR. *7200 CLA CLL TAD C7205 DCA SUBR TAD C7215 JMP I XTYPE TAD C7211 DCA SUBR TAD LF JMP I XTYPE TAD C7215 DCA SUBR TAD DOLLAR JMP I XTYPE DCA FLC TAD C7221 DCA SUBR JMP I XREAD RTR RTR DCA TEMP TAD LE DCA FLC TAD C7231 DCA SUBR JMP I XREAD BSW TAD TEMP DCA TEMP TAD C7237 DCA SUBR JMP I XREAD RTL RAL TAD TEMP DCA TEMP TAD C7246 DCA SUBR JMP I XREAD TAD TEMP DCA TEMP KSF JMP KRB TLS TSP 6'” o - l TAD MD SNA JHP I XSTORE TAD M3 SNA JMP I XEXEC TAD C30 SNA CLA JMP . +6 .-1 /START OF THE MONITOR /SET UP FOR 'JMS" IOUTPUT A CR /AND A LINE FEED /DOLLAR SIGN /INPUT AN OCTAL NUMBER /BITS 0-2 /INPUT /B1TS 3-5 /COMBINE /BITS 6-8 /BITS 9-11 /NOW GET CONTROL CHAR /SUBTRACT /IIDII 0? IYES; DEPOSIT /NO, '0"? /YES: BEGIN EXECUTION /N0' I/ll? in II 150 07267 1341 QM. TAD C277 /NO. TYPE '7' AND RESTART 07270 6046 TLS 07271 6041 TSF 07272 5271 JMP .-1 07273 5200 JMP MONITR 07274 1000 TAD TEMP /YES. DISPLAY CONTENTS 07275 3007 DCA PCSAVE /FROM MAINFRAME MEMORY 07276 1336 DSPLAY. TAD C7304 07277 3006 DCA SUBR 07300 1407 TAD I PCSAVE 07301 7006 RTL 07302 7006 RTL 07303 5747 JMP I XOUT /DISPLAY BITS 0-2 07304 1343 TAD C7311 07305 3006 DCA SUBR 07306 1407 TAD I PCSAVE 07307 7002 BSW 07310 5747 JMP I XOUT /DISPLAY BITS 3-5 07311 1344 TAD C7317 07312 3006 DCA SUBR 07313 I407 TAD I PCSAVE 07314 7012 RTR 07315 7010 RAR 07316 5747 JMP I XOUT /DISPLAY BITS 6-8 07317 1345 TAD C7323 07320 3006 DCA SUBR 07321 1407 TAD I PCSAVE 07322 5747 JMP I XOUT /DISPLAY BITS 9-11 07323 5200 JMP MONITR 07324 7477 XFLDCHG.FLDCHG 07325 7221 C7221. 7221 07326 7205 C7205. 7205 07327 7211 C7211. 7211 07330 7215 C7215. 7215 07331 0212 LF. 212 07332 0244 DOLLAR. 244 07333 7231 C7231. 7231 07334 7237 C7237. 7237 07335 7246 C7246. 7246 07336 7304 C7304. 7304 07337 7474 MD. -304 07340 7775 M3. -3 07341 0277 C277. 277 07342 0030 C30. 30 07343 7311 C7311. 7311 07344 7317 C7317. 7317 07345 7323 C7323. 7323 07346 7407 XREAD. READ 07347 7400 XOUT. OUT 07350 7472 XTYPE. TYPE 07351 7515 XSTORE. STORE 07352 7520 XEXEC. EXEC 7400 *7400 07400 0325 OUT. AND C0007 IBITS 9-11 ONLY 07401 1327 TAD C260 /MAKE IT ASCII 07402 6046 TLS 07403 6041 TSF 07404 5203 JMP .-1 07405 7300 CLA CLL 07406 5406 JMP I SUBR IRETURN 07407 6031 READ. KSF 07410 5207 JMP .-1 07411 6036 KRB /INPUT A CHARACTER 07412 6046 TLS /ECHO 07413 6041 TSF 07414 5213 JMP .- 07415 3017 DCA C 07416 07417 07420 07421 07422 1017 0330 1331 7510 5230 COMMAND? 07423 07424 07425 1332 7500 5230 COMMAND? 07426 07427 07430 07431 07432 07433 07434 07435 07436 07437 07440 07441 07442 07443 07444 07445 07446 07447 07450 07451 07452 07453 07454 07455 07456 07457 07460 07461 07462 07463 07464 07465 07466 1333 5406 7200 1016 7440 5736 1017 1322 7450 5735 1323 7440 5247 7240 1007 3007 5735 1324 7440 5256 2007 7000 7200 5735 1332 7450 5277 7200 1017 1340 7450 5270 7200 COMMAND 07467 07470 07471 07472 07473 07474 07475 07476 07477 07500 07501 07502 07503 07504 07505 07506 07507 07510 07511 07512 07513 07514 5736 1341 5737 6046 6041 5273 7300 5406 6031 5277 6036 6046 6041 5303 0325 7100 7004 7006 1326 3002 4001 5734 TYPE. FLDCHG. KSF JMP 151 CHAR C177 M60 0+6 M10 .+3 C10 I SUBR FLG I XOM CHAR M257 1 XDSPLAY M15 .+5 PCSAVE PCSAVE I XDSPLAY M2 0+5 PCSAVE I XDSPLAY M10 FLDCHG CHAR M224 .+3 I XOM C215 1 XMODEI .-1 CLL I SUBR 0.1 .-1 C0007 K5203 FLDSW FLDSWPI I XMONITR IJUST 7 BITS /LOWEST OCTAL NUMBER /NOT AN OCTAL NUMBER. IS IT A VALID /NOT AN OCTAL NUMBER. IS IT A VALID IRESTORE /RETURN /ONE OCTAL NUMBER MUST BE FOLLOWED IBY THREE MORE /18 IT A 'l'? /YES. DISPLAY CONTENTS /NO. IS IT A "<"? /YES. OPEN PRECEDING LOCATION /DISPLAY CONTENTS /NO. IS IT A '>'? /YES. OPEN NEXT LOCATION /DISPLAY CONTENTS INO. IS IT AN “F”? /YES. CHANGE FIELDS /NO. IS IT CONTROL-T? /YES. RETURN TO TRANSPARENT MODE /NO. TYPE QUESTION MARK AND IGNORE IPROMPT TO 05/8 MONITOR IOUTPUT A CHARACTER IGET FIELD NUMBER /ECHO IT /BITS 9-11 ONLY IFORM CDF CIF COMMAND /EXECUTE FIELD CHANGE 07515 07516 07517 07520 07521 07522 07523 07524 07525 07526 07527 07530 07531 07532 07533 07534 07535 07536 0753 07540 07541 07776 07777 1000 3407 5734 6001 5400 7521 7763 7776 0007 6203 0260 0177 7720 7770 0010 7200 7276 7267 7012 7554 0215 7776 7000 5776 152 STORE. TAD TEMP DCA I PCSAVE JMP I XMONITR EXEC. ION JMP I TEMP M257. -257 M15. '15 M2. '2 C0007. 7 K5203. 6203 C260. 260 C177. 177 ”60 9 -60 010. 10 XMONITR.MONITR XDSPLAY.DSPLAY XQM. QM XMODEI. POP-4 M24 9 -224 C215. 215 /GET THE DATA IDEPOSIT INTO MAINFRAME MEMORY /RESTART /REQUIRED TO LEAVE OF MEMORY /CHAIN TO MAINFRAME MEMORY / /ACTUAL CONTROL-PANEL INTERRUPT ENTRY POINT = 7777 / STRT. / $7776 MODEI JMP 1 STRT APPENDIX B MICROCOMPUTER CONTROL OF THE FLUORIMETER A. Introduction As discussed in Chapter V, the absorption-corrected spectro- fluorimeter constructed in this work for study of the cell shift method is controlled by an Intersil IM6100 microcomputer. Control is accomplished through three distinct electronic interfaces: a data acquisition interface; a cell positioner control interface; and a monochromator control interface. Both the design and operation of these circuits are described on the following pages. The reader is again referred to the PCM-lZ System Operation Manual (Pacific Cyber/Metrix, Inc., San Ramon, CA) for a deeper understanding of the diagrams and discussion presented. The hardware description is followed by a description of the microcomputer software developed to operate the interfaces and to transmit data to the minicomputer. A description of the minicomputer software developed to process the raw fluorescence data concludes this appendix. 153 154 B. Hardware Design and Operation l. Circuit Layout The microcomputer-fluorimeter interface circuits are constructed primarily on two printed circuit cards which plug into the IM6lOO data and control bus plane. A third interface card is also used, but contains only the radiation chopper monitoring circuitry. To simplify the interface circuits and to reduce the number of electronic components required, all I/O instruction decoding is performed on the interface card that contains the analog-to-digital converter (ADC). The decoded device select signals for the cell positioner and monochromator control circuits are fed from this card onto un- used lines of the IM6100 bus plane where they are picked up by the other interface cards. 2. Instruction Decoding Figures Bl and 82 illustrate the decoding circuitry for the entire microcomputer-fluorimeter interface. The identities of the electronic components shown are given in Table Bl. Data bus lines DXO-DXll are continuously buffered onto the ADC card by tri-state drivers U1 and U2. Early in the device I/O instruction cycle. the IM6100 CPU places the lZ-bit I/O instruction on these lines and pulses the LXMAR control line H1. The trailing edge of the LXMAR pulse latches the instruction into D flip-flops U7 and U8. A short time later, the DEVSEL line is pulsed L0 enabling devices U9-Ull to decode the nine least significant bits of the instruction. 155 lb ’4 1K in i”’ x xxx drum. 5E0" Agppzogwu czwuoowa cowpuzgpmcH o\H h a. F cu in 2a n HH— a «A «a. _ td a nu Guql mu? I m. N gun: a ta 2 n., o O“ m... x~.~ Du D“ n N. Ga .—m mgsupu no D. ,D o a n_nun_ 240 n m u c. a. '2 a: t n N: c 0— «a Nu m a o ¢49 ..u 0 _.~ . a. 9.— n+. m+ “.3 a u b. o a . n_nun_ uau , m m u c. a. t. h: r n a: t a. .. .wni o u o :40 - n, ,0— . .. [an—L n+ n+ ¢wo I\I\I\I\I\I\ I’GIOlv-dldl cncnoncncncu VVVVVV Amunv AAAAAA C'OICDC\‘OUO erOCOCOCOCO \I\I\I\I\I\I Aamuv 156 nonv at”. nbcv ..tv .55- «0“» “One “may .60. Ana. noun a... a... bun. Aguwzugwu pumpmm muw>mc mono tone .Nm mgaawu ohx Kano urn Arcv 157 Table Bl. Identity of Interface Circuit Components. Circuit Symbol Device Identity U1-U6, UE-UH, UK U7. U8 U9-U11 U12, U13 U14 U15 U16 U17 U18-U23 UA-UD UI UJ OAT-0A3, AMl-AM3 74L5365 74L5174 74L5138 Hex Tri-state Buffer Hex D Flip-Flop Decoder 06191 or IH5043 Dua1 SPDT FET SWITCH 74123 7474 Dual Monostable Multivibrator Dual D Flip-Flop Burr Brown ADCBOAG-lZ 12 Bit Analog-to- Digital Converter 74LSO4 7432 7473 7408 7417 Hex Inverter Quad 2-Input OR Gate Dual J-K Flip-FlOp Quad 2—Input AND Gate Hex Open Collector Buffer LF351 or CA3l40 BiMOS Op Amp 158 The three most significant bits are decoded within the CPU and used to generate the DEVSEL pulse. All I/O instructions that operate the,fluorimeter interfaces are of the form 65NM where N = 0, l or 2 and M = 0, l, ... 7. The instructions are fully decoded by the OR gates on the appropriate outputs of decoders U9-Ull as shown in Figure 82. Because the CPU executes the sequence described above twice during each I/O instruc- tion cycle, first with XTC exerted HI and then with XTC exerted LO, the signal 775 is ORed with the decoded device select signals so that the interfaces will be activated only once by each I/D instruction that is executed. All decoded device select signals are active L0. The instructions that operate the fluorimeter interfaces and the actions they initiate are summarized in Table 82. 3. Data Acquisition a. Signal Conditioning - Photocurrents from the fluorescence and reference detectors are converted to voltages, amplified, and filtered with the amplifier circuit shown in Figure 83. The circuit components are listed in Table Bl. Identical amplifiers are used in each signal channel. The input stage of the amplifiers is a simple current-to-voltage converter with a transfer function of lo6 V/A. The second stage is an inverting amplifier'with variable gain in decades from 10" to 104. The output stage is a unity gain inverting amplifier-first order active low pass filter with a variable time constant from 5 us to l50 ms. The overall transfer ‘function of the amplifiers is variable from 105 to 101° V/A in 159 Table 32. I/O Instructions for the Fluorimeter. Octal Code Action Initiated 6501 Set S/H to HOLD mode and start A-to-D conver51on 6502 Skip on end of A-to-D conversion 6503 Jam transfer ADC result into the accumulator 6504 Set multiplexer to reference channel 6505 Set multiplexer to fluorescence channel 6506 Set S/H to SAMPLE mode 6510 Skip on Y position encoder L0 6511 Skip on X position encoder L0 6512 Reverse power to Y motor ON/OFF 6513 Forward power to Y motor ON/OFF 6514 Forward power to X motor ON/OFF 6515 Reverse power to X motor ON/OFF 6516 Skip on Y origin encoder L0 6517 Skip on X origin encoder L0 6521 Set/Clear scan flip-flop 6522 Enable/Disable emission scan 6523 Enable/Disable excitation scan 6524 Skip on radiation chopper encoder L0 6525 Initiate control panel interrupt to enter PROM monitor 160 xgpwsuceu Lowewpcsq pcmgcauoposa .mm weaned l 1 i 81 "1 an an .ut and um flan um Bond ma non. v h: nun. u: no. n: mu. .m: n. .MS n.u TTTTTTTTTT 161 decades and is typically set at 107 V/A for the reference channel and at 109 V/A for the fluorescence channel. The potentiometer in the output stage can be used to offset the amplifier outputs con- tinuously from about -12 to +12 V to accommodate the input window of the analog-to-digital converter. Currently this window is -2.5 to +2.5 V. A time constant of 50 us is used in each signal channel to pass the 500 Hz modulated signals with a minimum of distortion. b. Sample-and-Hold and Multiplexinngircuitry - After amplifica- tion and filtering, the reference and fluorescence signals are passed to dual synchronously gated sample-and-hold (S/H) circuits and a two channel multiplexer shown in Figure B4. Initially, when power to the microcomputer is turned on, outputs 01 and 02 of flip-flop U15 are cleared. This sets the multiplexer to the reference channel and the S/H to the HOLD mode. To sample the reference and fluores- cence signals, a 6506 instruction is executed. This reverses the state of output 01 and closes FET switches l and 2 of device U13. Execution of a 6501 instruction clears output 01 of U15 to enter the HOLD mode. The 6501 instruction also triggers monostable U14 to issue a 700 ns logic HI pulse to the ADC to initiate analog-to- digital conversion. This pulse is issued after a 4 us delay to allow the S/H circuits to settle. The acquisition time of the S/H circuit is estimated to be about 1 us for a 5 V step signal to an accuracy of about one percent. Since the average device I/O instruc- tion time of the IM6100 CPU is 8.5 us, the speed of these circuits is more than sufficient for use with the microcomputer. 162 #20 um: baa an x3585 $582.53: En Boxiccmim—QEmm .em aeseee O O I, O b¢u>zou baa S\¢ 163 Multiplexing of the S/H outputs to the ADC is accomplished with FET switches l and 3 of device U12. Execution of a 6506 instruction forces output 02 of flip-flop U15 HI which opens switch 3 and closes switch 1 to pass the fluorescence signal. Execution of a 6504 instruction clears output 02 of U15 and reverses the states of the switches to pass the reference signal. Follower 0A1 is necessary to prevent an IR voltage drop across the switches. The settling time of the multiplexer is about 1 us, and so, poses no problem for use with the microcomputer. c. Analog-to-Diqital Conversion - Figure BS shows the circuit used for analog-to-digital conversion of the signals from the fluorimeter. After receiving the CONVERT pulse from monostable U14, the ADC (U16) requires 25 us to complete the digital conversion. At the end of the conversion, the EOC output is exerted LO. After starting the conversion, the microcomputer is programmed to execute 6502 instructions in a program loop to test for the end of conversion. The first 6502 instruction executed after EOC goes L0 enables tri- state buffer U6 to exert the SKP line L0 causing the CPU to exit from the end of conversion test loop. The result of the conversion is read by the CPU when a 6503 instruction is executed. Buffers U3 and U4 are enabled by this instruction to drive the contents of the ADC outputs into the 0X lines, and buffer U5 exerts control lines CO and Cl L0 and line 02 HI to cause the contents of the 0X lines to be jam transferred into the IM6100 accumulator. Because the reference and fluorescence signals from the 164 mxxu h¢u>zcu baa 1\m uhx xonm mcupzucwu cowmgm>cou.quwcvoicuico_mc< .mm aeseee mum away «no 000 AAA Ole-IO ummn VVV AAAAAA tmflnam 661501610] vvvvvv AAAAAA cannula rmommm vvvvvv 165 fluorimeter are modulated, the microcomputer must synchronize the analog-to-digital conversions with the modulation cycle of the optical chopper. The circuit used to monitor the chopper wheel is shown in Figure B6. When the optical beam of the GEHlBBl detector is blocked by a blade of the chopper wheel, pin 1 of tri-state buffer UK is exerted L0. To test for this condition, the microcomputer is placed in a program loop repetitively executing 6524 instructions. The first 6524 instruction executed after the chopper detector is blocked enables UK to drive the SKP line LO causing the microcomputer to exit from the test loop. Through an appropriate software routine, the same circuit is used to detect when the chopper detector is not blocked by a chopper wheel blade. 6 Because the signal from the chopper detector and the fluores- cence and reference signals may not be exactly in phase due to the position of the chapper detector, a timing delay is employed between the change of state of the chopper signal and the initiation of data acquisition. A delay of about 300 s was found to be suitable in the present instrument and is accomplished by a software timing 100p. 4. Cell Positioner Control Circuitry a. Cell Position Monitoring - The circuitry used to monitor the position and movement of the fluorescence sample cell is also shown in Figure 86. The circuitry and its operation are identical to that associated with the radiation chopper wheel. Four GEHl3Bl detectors are located on the cell positioner, two associated with encoder wheels on the motor shafts, and two located to define an origin 166 C, “JGW‘N’ vvvv V A AAAA H \IUIU'IU (11) (16) (25) (5‘1) CHOPPER Figure B6. Cell Position and Chopper Hheel Monitoring Circuitry 167 in the x-y plane. When the optical beam of one of these detectors is blocked by an encoder wheel blade or by a vane on the positioner, the enable input 51 of the corresponding tri-state buffer UE, UF, US, or UH is exerted LO.. The microcomputer checks for this condition by executing skip test instructions in a program loop. The 6510 and 6511 instructions test the states of the y and x encoder wheels respectively. The y and x origin encoders are tested by 6516 and 6517 instructions. When one of these instructions is executed, enable input 52 of the corresponding buffer is exerted LO. If 61 is LO at that time, the buffer drives the SKP line L0 causing the microcomputer to jump out of the test loop. The displacement of the cell during a movement is determined simply by counting the number of LO-HI and HI-LO logic transitions from the position encoders. b. Control of Power to the Cell Positioner - Power is supplied to the cell positioner from a separate 12 V DC power supply by the circuit shown in Figure B7. When the IM6100 is first turned on and the R/H bus line is pulsed LO by the PROM board circuitry, flip- flops UA and U8 are cleared. This turns transitors 01, Q3, 05, and 07 ON while 02, Q4, Q6, and 08 are turned OFF. Both sides of each motor on the positioner are at the same voltage (about 9.0 V) and, therefore, no current flows through the motors. To move the cell in the +y direction, a 6513 instruction is executed. This toggles flip-flop Al and turns transitors 01 OFF and 02 ON. One side of the y motor is grounded causing current to flow through the motor. Execution of a second 6513 instruction reverses the state of flip- flop Al and stops the current flow. Movement in the -y direction 168 +12 <73) +12 v t 8 OJ 0 - o 1 .21? 02 6513 <12) ____ wv (1“) 6512 v *12 uE>Ji - 03 +3 $55 04 n1 <21) R’H‘ ’ y ‘P 1 2 JUD5;2 [1;Eig§$°s .1111. §. 8514 (26) .... wx 6515 + <23) . van :2 5E>£T - 07 as flit Figure B7. Cell Positioner Control Circuitry. R1 = 2.2K. R2-R5 = VR1 - VR4 = 1K, 01 - 08 = 2N6043. 169 is accomplished by execution of a 6512 instruction. This instruc- tion toggles flip-flop Bl and turns transistors 03 OFF and 04 ON. The opposite side of the motor is grounded by this action causing the current and the direction of motor shaft rotation to be reversed. A second 6512 instruction switches the current off. The x motor is controlled in an identical manner by 6514 and 6515 instructions. c. Control Sequence (i) Movement to the Origin - To move the fluorescence cell precisely to any point in the x—y plane, it must first be positioned precisely at a reference point. Movement to the x-y origin is ac- complished by first applying a DC current to the x motor to move the cell rapidly toward the x coordinate of the origin. When the origin detector beam is blocked by the vane on the positioner, the current is switched off and the motor is allowed to coast to a stop. The same procedure is then executed in the y direction. Next, a 5 Hz pulsed DC current of opposite polarity is applied to the x motor to move the cell slowly away from the origin detector. When the beam of the origin detector becomes unblocked, the pulse train is con- tinued until a HI-to-LO logic transition is observed at the x en- coder wheel. This sequence is then repeated for the y motor. In this manner, the cell approaches all points in x-y space, including the origin, from the same x and y directions. This greatly reduces errors due to the looseness of the dovetail slides and the lead screw threads. Also, the encoder wheels are actually used to define the origin as well as the coordinates of all other 170 points in the x-y plane. This ensures that all cell positions are highly reproducible. To discriminate against switching transients picked up from the motor power lines, the signals from the origin and position detectors in the sequence described above are always double checked after a 1 ms delay. (ii) Movement to Other Cell Positions - To move the cell to other positions in the x-y plane, a series of 2.5 ms current pulses of gradually decreasing frequency is applied to the motors. This is accomplished in a three part sequence. The pulsed DC current is first applied at about 100 Hz until the motor shaft is one revolution short of its total rotation for the movement. The pulse frequency is then reduced to about 40 Hz for 0.9 revolution. Finally, current pulses are applied at about 5 Hz until the last encoder signal is counted. During the final part of this sequence, the motor shaft advances about 0.02 revolution with each current pulse and the cell moves about 0.001 cm. If the cell movement involves a lead screw rotation of one revolu- tion or less, the first part of the sequence is skipped. If a rota- tion of only 0.1 revolution is involved, only the final part of the sequence is used. For the work described in this thesis, all cell movements between cell positions and between the origin and cell position 1 involved equal movements of 130 encoder logic transitions or 0.65 cm in both the x and y dimensions. 171 5. Monochromator Control Circuitry - To scan the GCA McPherson monochromators used in the fluorimeter, a TTL square wave is gen- erated under microcomputer control by the circuit shown in Figure B8 and used to drive the external scanning inputs on the monochromator slew boxes. With the slew box scan switches set to 20 A/s, a 1 nm scan is accomplished by generating 300 cycles of a 600 Hz square wave. The microcomputer does this by executing 6521 instructions at approximately 800 ps intervals to toggle flip-flop UDl. The square wave is then multiplexed to either or both the excitation and emis- sion monochromators by the action of flip-flops U01 and UC2 and gate UI. All flip-flops are initially cleared on power-up so that neither monochromator is scan enabled. Execution of a 6523 instruc- tion toggles flip-flop UCl allowing the square wave from UD to pass gate UI to the excitation monochromator. A second 6523 instruction reverses the state of UCl and disables the scan. Similarly, the 6522 instruction enables and disables the emission monochromator for scanning. Different wavelength intervals can be scanned by simply generating a different number of cycles of the 600 Hz square wave. C. Software Structure 1. Fluorimeter Control and Data Collection A single PAL8 program named DRC3 was written for the IM6100 microcomputer to control all fluorimeter functions. The program is subroutine oriented around a command decoder which calls the various subroutines to execute the fluorimeter control functions. 172 q Figure BB. Monochromator Scanning Circuitry lJI l1! 5?... EH "ONO f” 172 z lJI q 5 III tn uowo J77 Figure BB. Monochromator Scanning Circuitry 173 All keyboard commands to DRC3 are single character commands, some of which require an octal argument. Table B3 sunmarizes the conmands and the actions they initiate. The binary program DRC3.BN is loaded into the microcomputer memory by issuing the following command to the 05/8 keyboard monitor: _-_ PU DRC3.BN/B Loading is executed automatically, and control of the keyboard is turned over to the IM6100 PROM monitor. To start DRC3, one simply types 3 F0 § 02006 Program DRC3 should respond by typing a "/" on the CRT to indicate that it is running and waiting to receive a conmand. 2. Data Transfer Several programs were written in OS/8 Fortran II-SABR for the PDP 8/e to receive, store, and correct fluorescence data collected by program DRC 3. The first of these programs is RCVR2.FT which receives 24-bit integer data transmitted from the IM6100 as fluores- cence-reference intensity pairs, converts the numbers to floating point format, normalizes the fluorescence to the reference intensity, and stores the results under a filename specified by the user. RCVR2 operates on a "handshaking" principle with DRC3 during data transfer. 174 Table B3. Keyboard Commands for Program DRC3. Command Resulting Action A Executes alignment routine after querying for NNNN8 points to take in the cell profile. Direc- tion of movement and displacement increment should be set prior to command. B Measure baselines in each signal channel and type results on CRT (chopper must be on). 0 Measure net fluorescence and reference signals and type results on CRT (chopper must be on). E Exit to transparent mode G Start scan LNNNN Set scan length to NNN8 nanometers. M Execute cell movement (direction and displacement must be set prior to command). 0 Move cell to the x-y origin. R Collects data at cell positions 1, 2, and 3. Sequence is repeated NNN8 times after querying for this argument. SN Set up for scan type N where N = l) for excitation scan 2; for emission scan 3 for synchronous scan T Transmit data to minicomputer (program RCVR2 must be loaded and started on the 8/e prior to conmand). XMNNN Set up for X cell displacement of NNNg steps in the M direction, M = F(forward) or R(reverse). One step = 0.005 cm. YMNNN Set up for Y cell displacement of NNN8 steps in the M direction. 175 When a data transfer is to be made, the user exits from DRC3 to the 05/8 monitor and loads and starts RCVR2. The program responds by typing "Ready" on the CRT and then waits for an octal 200 character to be transmitted from the microcomputer to signal the beginning of data transfer. The user then restarts DRC3 and types "T" to initiate data transfer. Program DRC3 immediately sends an octal 200 to the 8/e and then transmits data characters in response to octal 100 characters which are sent to the microcomputer by RCVR2 when it is ready to receive a character. Data transfer is terminated when DRC3 sends a second octal 200 character after the last data character. 3. Data Correction Corrections for primary and secondary absorption errors, reflec- tion effects, and the detection system sensitivity are coordinated by program FLUOR0.FT. The corrections are executed according to a list of correction parameters and a list of file specifications supplied to the program by the user. The correction parameters include the fluorimeter window parameters and cell reflectance values. The file specifications include input and output file names, file length, and file type. These specifications are then passed to the correc- tion program STAT.FT if the user has specified static wavelength data, or to the program SPEC.FT if spectral data has been specified. Both of these programs are swapped into core by the Fortran II chain- ing feature and, in turn, chain back to program FLUORO when execution is finished. More specific information about these programs can be found in the program listings. 176 4. Program_Ljstings On the fOllowing pages, listings of the source files of programs DRC3.PA, RCVR2.FT, FLUORO.FT, STAT.FT, and SPEC.FT are presented. Some important subroutines are also listed immediately following the Fortran programs in which they are referenced. These include sub- routine SUBIN.FT which handles keyboard input for FLUORO.FT, sub- routine SUBCORR.FT which performs the absorption and reflection correction calculations for STAT.FT and SPEC.FT, and subroutine SD.FT which computes sample mean and standard deviation values for program STAT.FT. 177 /////////////////////////////////////////// \‘\\‘\\ ////////////////////////////////////////// CALLSJIS I 0 PROGRAM DRC3 . PA FLUORIMETER CONTROL ROUTINE / / / / / / /INDIRECT SUBROUTINE CALL /SET S/H TO SAMPLE MODE ISET S/H TO HOLD MODE AND START ADC /JAM TRANSFER ADC RESULT INTO AC /SKIP ON A-TO-D DONE FLAG /SET‘ MULTIPLEXER TO /SET MULTIPLEXER TO F CHANNEL [CHANNEL IFORWARD POWER TO X MOTOR 0N IFORWARD POWER TO X MOTOR OFF IREVERSE POWER TO X MOTOR ON IREVERSE POWER TO X MOTOR OFF /FORWARD POWER TO Y MOTOR ON IFORWARD POWER TO Y MOTOR OFF /REVERSE POWER TO Y MOTOR ON lREVERSE POWER TO Y MOTOR OFF /SKIP 0N X ENCODER LOW ISKIP ON Y ENCODER LOW ISKIP 0N X ORIGIN LOW ISKIP ON Y ORIGIN LOW /SKIP ON CHOPPER LOW /COMMANDS FOR SERIAL PORT TO B/E /RETUR.N TO PROM MONITOR IFREQUENT'LY ADRESSED REGISTERS 5MPL=6506 ADST*6501 ADBB=6503 ADSF=6502 HUXI=6504 HUXF=6505 XFON365I4 XFOF=XFON xn0N=6515 XROF=XRON YFON=65I3 YFOF‘YFON YRON=6512 YBDF‘YRON SXEL=6511 SYEL=65IO SXOL=65I7 SYOL=6516 SCHL=6524 PDPTLS=6146 PDPTSF=6141 PDPKSF=613I PDPKBB=OI36 EX1T=6525 FIELD 0 *10 TDATA. 0 POINTERqO $20 NOPTS. 0000 BB. 0 LB, 0 "DIV. 0 LI. 0 HI. 0 LP. 0 BF, 0 TEMP. 0 DTIME. 0 HIB. 0 LIB. 0 HFB, 0 ”B. o SBFLAC. 0 COUNT. 7777 TIME. 0 NOVAL. O FLAGI, 0 n1. -1 m. -2 M3. '3 H4. -4 M5. '5 M10. '10 M11. “II MI2. '12 M14. ‘14 M15. '15 /POINTER.TO DATA STORAGE lMINUS THE NUMBER OF ADC‘S TO AVERAGE IHIGH’BYTE 0F DIVIDEND /LOW BYTE 0F DIVIDEND IMINUS THE VALUE OF THE DIVISOR /LOW’SUMMATION REGISTER FOR I IHIGH SUMMATION REGISTER FOR I /LOW SUMMATION REGISTER FOR F /HIGH SUMMATION REGISTER FOR F /TEMPORARY CHARACTER STORAGE /NO. OF DELAY TIMING CYCLES /BASELINE SUMMATION REGISTERS ISIGNAL-BASELINE FLAG INUI‘IBER OF ENCODER PULSES TO COUNT IDELAY TIME /NO. OF DATA POINTS IVARIOUS CONSTANTS MON, HBO, M22, M30, M50, M60, M76, M100, M1 15. M144, mozg M454, M1750. P3, P10, P11, P77, P177. P200, P256, P260, P6510. P6511. P6512. P6513, P6514. P6515. P6777. ADCNTI. ADCNT2, ADCNT3. DIVIDE. STRING. DECOUT. -20 ~22 -3O -50 -6O -76 -100 -115 -144 '202 -454 -1750 10 11 77 177 200 256 260 6510 6511 6512 6513 6514 6515 6777 O O O 178 IPOINTERS TO SUBROUTINES DIV STR ARGU RORF LEN SPEC MODE2 SCN ERR RING /DIVISION SUBROUTINE. ENTER WITH DIVIDEND IIN HB-LB AND DIVISOR IN MDIV /TYPE OUT ASCII STRING IMMEDIATELY /FOLLOWING THE CALL COMMAND ITYPE OCTAL VALUE 1N AC AS A DECIMAL NO. /TYPE ASCII CHARACTER IN THE AC /TYPE A CARRIAGE RETURN AND LINE FEED IINPUT AN OCTAL NUMBER IINPUT AN ASCII CHARACTER /DATA ACQUISITION SUBROUTINE lBASELINE ACQUISITION SUBROUTINE /CELL TRANSLATION SUBROUTINE ICOMMAND MONITER /PREPARE FOR MOVEMENT IN XIDIRECTION IPREPARE FOR MOVEMENT IN Y'DIRECTION /RETURN TO XfY ORIGIN /TIME DELAY SUBROUTINE /SEND DATA TO B/E /GET 4 DIGIT OCTAL NUMBER ISTORE BASELINE CORRECTED DATA / DISPLAY ADC RESULTS IALIGNMMENT ROUTINE ITHREE POINT ACQUISITION ROUTINE /APPLY 2.5 MS CURRENT PULSE TO MOTOR ISUBR.TO ESTABLISH MOTOR.DIRECTION /READ IN 3 DIGIT OCTAL ARGUMENT, RETURN NEG IN AC IDETERMINE MOTOR DIRECTION (FOR OR.REV) /SET SCAN LENGTH ISTART SCAN ISET SCAN MODE ISCAN MONOCHROMATORS IERROR MESSAGE FOR.POSITIONING ERROR IRING THE BELL ON THE'FTY ICOMMAND MONITOR 179 DCA FLAG] CALL CRLF TAD (25? CALL TYPE ITYPES A 'l' WHEN READY FOR INPUT CALL IN /GET SINGLE CHARACTER.COHMAND TAD (-301 SZA l'A'? JMP .+3 CALL ALIGN JMP HON TAD H1 82A /'B'? J” 0 +4 CALL BASE CALL OUT JMP HON TAD M2 SZA /'D'? JMP .+4 CALL DATA CALL OUT JMP HON TAD Ml SZA I'E"? JMP .+2 EXIT TAD M2 SZA /'G"? JMP .+3 CALL GOGET JMP MON TAD H5 SZA /'L'? JMP .+3 CALL LENGTH JMP MON TAD Ml SZA /'M"? JMP .+3 CALL MOVE JMP MON TAD n2 SZA /'O'? JMP .+4 CALL ORIGIN CALL BELL JMP HON TAD HO SZA /'RF? JMP .+3 CALL BUN JMP MON TAD n1 SZA /'S'? JMP .+3 CALL SCHODE JMP HON TAD n1 SZA /'T'? JMP .+3 CALL SEND JMP HON TAD H4 SZA /'X'? JMP .+3 CALL XSET JMP HON ENCI, SHORT'3 . SHORT 1 . SHORT2 . PULSE . ENC2 . ENC3 . ENCG , ENC5 . ENC6 . XS. 180 TAD Ml SZA JMP .+2 CALL YSET JMP MON /NOT A LEGAL COMMAND. ‘00 /CELL TRANSLATION SUBROUTINES : 0 /MOVE THE CELL CALL STEP TAD COUNT IAC SNA JMP SHORTl 0 /POSITION ENCODER MUST BE HIGH JMP .+2 /AT' START OF MOVEMENT CALL ERROR TAD P11 SMA JMP SHORT2 DCA CNTRM TAD TIME DCA DTIME JMS G02 TAD M11 DCA CNT'RM TAD M76 DCA DTIME JMS 002 TAD M1 DCA CNTRM TAD M144 DCA DTIME JMS G02 JMP I MOV SNA JMP SHORT3 TAD M11 JMP SHORT‘3+1 O JMS PULSE ISZ CNTRM JMP .-2 CLA JMP I 002 0 : /ROTATE LEAD SCREW 18 DEGREES JMP .+12 CALL STEP CALL DELAY 0 JMP .+2 JMP .-4 0 JMP I PULSE JMP .-7 CALL STEP CALL DELAY 0 JMP 0 JMP .-5 JMP I PULSE 0 /SET‘ UP FOR X TRANSLATION CLA TAD M10 DCA TIME /"Y"? IGNORE IT! .-3 XFOR. YFOR, ENC. ORG, 181 TAD P6511 JMS ENC TAD FLAGI SZA CLA CALL FR JMP XFOR CLA TAD P6515 CALL DIRECT CLA CALL ARC DCA COUNT JMP 1 XS CLA TAD P6514 CALL DIRECT CLA TAD FLAG! SNA CLA JMP .+3 CALL ARC DCA COUNT JMP I XS 0 /SET UP FOR Y TRANSLATION CLA TAD M5 DCA TIME TAD P6510 JMS ENC TAD FLAGI SZA CLA CALL FR JMP YFOR CLA TAD P6512 CALL DIRECT CLA CALL ARC DCA COUNT JMP I YS CLA TAD P6513 CALL DIRECT CLA TAD FLAGI SNA CLA JMP .+3 CALL ARC DCA COUNT JMP I YS 0 /SET UP DIRECTION OF MOVEMENT DCA ENC] TAD ENCI DCA ENC2 TAD ENC2 DCA ENC3 TAD ENC3 DCA ENC4 TAD ENC4 DCA ENC5 TAD ENC5 DCA ENC6 JMP I ENC 0 /RETURN TO ORIGIN CLA TAD M144 STEPR. DCA DTIME TAD P6514 JPIS DIR JMS STEPR CALL DELAY SXOL JP? .‘1'2 JP? .‘4 SXOL JP? .‘1'2 JP? 0.? TAD P1144 DCA DTIME SXEL JMP .+2 JP? .‘1'4 JMS STEPR CALL DELAY JP? .-5 JMS STEPR CALL DELAY SXEL JP? .+2 JP? 0-4 TAD P150 DCA DTIME TAD P6513 JPIS DIR JMS STEPR CALL DELAY SYOL JP? .‘1'2 JP? 0" SYOL JP? .+2 JP? .-7 TAD M100 DCA DTIPIE SYEL JP? .+2 JP? .+4 JMS STEPR CALL DELAY JP? .‘5 JMS STEPR CALL DELAY SYEL JP? .‘1'2 JP? 0-4 JP? I ORG 0 182 mm CHECK X ORIGIN ENCODER /DOUBLE CHECK Y ORIGIN ENCODER IANOTHER DOUBLE CHECK IANOTHER DOUBLE CHECK IGENERATE CURRENT PUISES ON, OFF, DIR, OWT, DIV, BEGIN. CLA TAD (7500 DCA CNTRP 0 I82 CNTRP JMP 0.1 0 JMP I STEPR 0 0 DCA 0N TAD 0N DCA OFF JMP I DIR 0 CALL CRLF CALL STRING 0215 0306 0275 0240 0000 TAD HF CALL DECODT TAD P256 CALL TYPE TAD LF CALL DECOUT CALL CRLF CALL STRING 0311 0275 0240 0000 TAD HI CALL DECOUT TAD P256 CALL TYPE TAD LI CALL DECOUT JMP I OWT 183 ITO THE APPROPRIATE ROTOR ISELECT THE ROTOR / OUTPUT RESULTS IOUTTUT F /OUTPUT I /UTILITY SUBRDUTINES 81000 0 CLA CLL TAD H15 DCA CNTRD CLA CLL TAD BB TAD HDIV SZL DCA BB JMS ROT ISZ CNTRD JMP BEGIN CLA JMP I DIV 0 0 CLA TAD RAL DCA TAD RAL DCA 5355 IDIVISION SUBROUTINE. RB LB ' DIVIDEND IHDIV 8 -DIVISOR.... BBzREfllINDERwLB=QUOTIENT STR, XIN. XREAD, CR, ARGU , JI‘IPIRDT TADISTR SPA SNA STR JMP STR+1 SPA JMP I TAD P10 JMP 1 CLA DCATEI‘IP CALLREAD RTL TAD TEMP DCA TEMP CALL READ TAD TEMP CIA JMP I ARGU CALL IN TAD (-306 SNA JMP I RORF TAD M14 SNA JMP .+2 JMP I I‘IONITR ISZ RORF 184 ITYPE TEE ASCII STRING IHHEDIATELY IFOLLOWING THE CALL COPIHAND ME OUT ASCII CHARACTER IN THE AC /R.EAD ASCII CHARACTER INTO AC IREAD OCTAL NUMBER INTO AC ITYPE CR LF /READ 3 DIGIT OCTAL ARGUMENT IREAD MR DIRECTION DEC. BAS, DLA. DCNTRI. DCNTR2, SAV, NOP JHP I RORF #1200 0 DCA TEHP TAD H1750 DCA HDIV TAD TEHP DCA LB DCA EB CALL DIVIDE TAD LB TAD P260 CALL TYPE TAD BB RAR CLL DCA LB DCA BB TAD H144 DCA HDIV CALL DIVIDE TAD LB TAD P260 CALL TYPE TAD HB RAR CLL DCA LB DCA EB TAD H12 DCA HDTV CALL DIVIDE TAD LB TAD P260 CALL TYPE TAD BB RAR CLL TAD P260 CALL TYPE JHP I DEC 0 CALL DATA TAD BIB DCA HI TAD LIB DCA LI TAD RFD DCA HF TAD LFB DCA LF JHP I BAS 0 CLA TAD DTIHE DCA DCNTR2 CLA TAD H115 DCA DCNTRI ISZ DCNTRl JMP .-l 182 DCNTRZ JHP .- JHP I DLA 0 0 0 TAD BI JMS STORE 185 ITYPE OCTAL NO. IN AC AS DECIHHAL /DIVIDE BY 1000 /OUTPUT THE INTEGER.PART /DIVIDE REHAINDER.BY 100 /OUTPUT THE INTEGER.PART /DIVIDE REHAINDER BY 10 /0UTPUT THE INTEGER PART /HEASURE BASELINE /TIHING LOOP ISTORE DATA STORE. ALINE. AGAIN, 8ND. A. 186 TAD LI JMS STORE TAD HF JRS STORE TAD LF JRS STORE JnP I SAV 0 /STORE IR FIELD l CDF 10 DCA I POINTER GDP 0 JR? I STORE 0 IALIGNHENT ROUTINE CALL CRLF CALL STRING 0316 0317 0256 0240 0320 0324 0323 0256 0240 0275 0240 0000 CALL RUHIN DCA ROVAL TAD NOVAL CIA DCA CNTRA TAD P177 DCA POINTER CALL DATA CALL SAVE CALL MOVE ISZ CNTRA JMP AGAIN CALL BELL JMP I ALIIIE 0 31400 0 ITRARSRIT DATA TO THE RINI TAD P177 DCA POINTER TAD NOVAL CLL RTL DCA CNTRT TAD P200 JMS PDPOUT JHS READY GDP 10 I POINTER GDP 0 DCA TEHP TAD TEHP /NO. POINTS = AND P77' JHS PDPOUT JMS READY TAD TEMP AND P77 JHS PDPOUT ISZ CNTRT CNTRT, READY. PDPOUT. PDPIN. RDNN. JMP A TAD P200 JMS PDPOUT JMP I SND JMS PDPIN AND P177 TAD M100 SZA JMP .-4 JMP I READY 0 PDPTLS PDPTSF J“? 0.1 CLA JMP I PDPOUT 0 PDPKSF JMP .-l CLA PDPKRB JMP I PDPIN 0 CALL READ RTR RTR DCA TEMP CALL READ BSW TAD TEMP DCA TEMP CALL READ RTL RAL TAD TEMP DCA TEMP CALL READ TAD TEMP JMP I NUM $1600 0 CLA DCA FLAG] TAD M202 DCA COUNT TAD P177 DCA POINTER DCA NOVAL CALL CRLF CALL STRING 'N 'O .0 'R 'U 'N 'S '3 n 0000 CALL ARG DCA RCNTR 187 /SEE IF MINI IS READY FOR.DATA ISEND CHARACTER 'IO MINI IREAD CHARACTER FROM'MINI /READ 4 DIGIT OCTAL ARGUMENT IDATA ACQUISITION SEQUENCE R1. RCNTR. MODE2. CHAR. SET. EXSET. EMSET. SYNSET. LEN, CNTRS. SPEC. CALL CRLF CALL YSET CALL MOVE CALL XSET CALL MOVE CALL DATA CALL SAVE ISZ NOVAL CALL OUT CALL MOVE CALL DATA CALL SAVE ISZ NOVAL CALL OUT CALL YSET CALL MOVE CALL DATA CALL SAVE ISZ NOVAL CALL OUT CALL ORIGIN ISZ RCNTR JMP R1 CALL BELL JMP I RUNN 0 0 CALL READ DCA CHAR JMS SET JMP I MODE2 0 0 TAD CHAR TAD M1 6523 JMP I SET 0 CALL ARC DCA CNTRS TAD CNTRS CIA DCA NOVAL JMP I LEN 0 0 CLA TAD P177 DCA POINTER CALL DATA CALL SAVE CALL SCAN ISZ CNTRS 188 /READ SCAN MODE /SET SCAN MODE /READ AND SET SCSAN LENGTH /SCANNING SEQUENCE SCN. CNT, WAIT. CNTRN. OK, JMP SPEC+4 CLA JMS SET CALL BELL JMP I SPEC CLA TAD M454 DCA CNT 6521 JMS WAIT 6521 JMS WAIT ISZ CNT JMP i scn CLA TAD M76 DCA CNTRW ISZ CNTRW JMP .- JMP I WAIT 0 $2000 CLA CLL SMTL TAD NOPTS JMP .+2 TAD LIB 189 /STEP TEE MONOCRROMATORS /T1MJNG LOOP FOR.STEPPING PULSES /A-TO-D SEQUENCE /INITIALIZE SUMMATION REGISTERS /DELAY TIME /SET FLAG FOR DARK SIGNAL /LOOK FOR DARK SIGNAL /TAKE DATA ISET FLAG FOR.LIGHT SIGNAL ITAKE DATA ICOUNT NO. OF POINTS ICOMPUTE NET SIGNALS GET. SIG. RIB HI LFB RFD I ADC P6777 TDATA I TDATA ADCNT3 SIG I GET 190 /A-TO-D CONVEIB ION IPOINTER '1!) DATA STORAGE /SET MUX TO REFERENCE CHANNEL ICONVERT ISET MUX TO FLUORESCENCE CHANNEL /CONVERT /STORE I TEMPORARILY /STORE F TEMPORARILY IDONE? /GO TO INTEGRATION ROUTINES /INTEGRATE LIGHT SIGNALS 191 BASL. CLL /INTEGRATE DARK SIGNALS TAD I TDATA CMA TAD LIB DCA LIB SZL ISZ RIB NOP CLL TAD I TDATA CMA TAD LFB DCA LFB SZL ISZ RFB NOP ISZ ADCNT3 JMP BASL JMP I GET $2200 ERR. 0 /TYPE ERROR MESSAGE ON CRT CALL CRLF CALL STRING 0000 JMP I MONITR RING, O /RING THE BELL CLA TAD P207 CALL TYPE JMP I RING P207. 207 8888 192 CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C PROGRAM RECEIVER FILE: RCVR2." C C C PURPOSE: RECEIVE AND STORE DATA C C C C 00 FROM THE [$100 MICROCOMPUTER CCCCCCCCCCCCCCCCCCCCCCCCCCCCGCCCCCCCCGCCCCCCC OPDEF BSW 7002 DIMENSION I(1000).F1(99).F2(99) .F3(99) VRITE( 1. 1) FORMAT(/’HELLO’/) READ( 1.5)FNAME FORMAT( 'NHAT IS THE FILENAME?(A6) 'A6) CALL OOPEIN ’FLP2‘ .FNAME) READ( 1. 10)L 10 FORMAT( ‘ IS THIS A SPECTRUM’NY/N) 'Al) IF( L- 1632) 15.250. 15 15 READ( 1.20) NSETS 20 FORMAT( ’HOW MANY 3 PT. DATA SETS IN THIS FILE‘N I2) ' I2) WRITE( 1.25) 25 FORMAT( 'OK! TYPE CONTROL-A TO ENTER LOCAL MDE. ’) S JMS DATA MK=1 DO 100 K=I,NSET‘S PM K) =FLOAT( I(MI(+2) )+2048.+( FLOAT( ”MK-+3) )+2048. )/4096. F1(K)=F1(K)/(FLOAT(1(MK))+2048.+(FLOAT(I(MK+1))+2048.)/4096. ) F2( K) =FLOAT( I(MK+6) )+2048. +( FLOAT( I(MK+7) )+2048. )/4096 . F2( K) =F2( K) /( FLOAT( I( MIG-4) )+2048. +( FLOAT( I( MIG-5) )+2048. )/4096.) F3( 10 =FLOAT( I(MK+10) )+2048. +(FLOAT( I(MK+11) )+2048.)/4096. F3( K) =F3( K) /( FLOAT( I( MK+8) )+2048. +( FLOAT( I( MIG-9) )+2048. )l4096 .) MK=MK+12 100 CONTINUE DO 150 K=1.NSETS WRITE(4.110)F1(K) .F2(IO .F3(K) 110 FORMAT( 3E15.9) 150 CONTINUE 175 CALL OCLOSE READ( 1 .200)L 200 FORMAT( 'DO YOU HAVE ANOTHER FILE?( Y/N) ’Al) IF( L-1632)400.2.400 250 WRITE( 1.25) S JMS DATA X80. DO 300 N3 1 .J.4 N1=N+1 N2=N+2 N3=N+3 F1( 1)=FLOAT( 1(N) )+2048.+(FLOAT( I( N1) )+2048. )l4096. F1( 1)‘-'(FLOAT( 1(N2) )+2048.+(FLOAT( I(N3))+2048. )/4096.)/F1( 1) X=X+1. WRITE(4.275)X.F1( 1) 275 FORMAT( ’RD' .2E15.5) 300 CONTINUE GO TO 175 400 STOP WOOOOOO ONI- mmmmmmmmmmmmmmmmm 193 JMS PROMPT JMS IN JMP EOF TAD TEMP TAD (4000 DCA m J=J+1 I(J)=M JMP GET 0 CLA JMP I DATA TAD (-20O JMP I IN 182 IN AND (77 JMP I IN TAD (100 CLA JMP I PROMPT 194 CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C C PROGRAM FLUORO.FT “C C C C FLUORESCENCE CORRECTION C C COORDINATOR C C C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC COMMON RH01,RHO2.W1.W2.W3.W4.TH1.TH2.TH3.TH4 COMMON IXMIN.NOPTS.INTVAL.K.FILE1.FILE2.FILE3.LTYPE C C BEGINNING OF COMMAND MODE C WRITE(1.1) 1 FORMAT(/6X5’PROGRAM FLUORO VERSION 1’/) 2 CALL IN(L) C C COMMAND DECODER C 5 IF(L-513)10.170.10 10 IF(L-344)11,235.11 11 IF(L-333)12.236.12 12 IF(L-l241)15,237,15 15 IF(L-1236)20.240.20 20 IF(L-784)25.189.25 25 IF(L-774)30.194.30 30 IF(L-463)35.200.35 35 IF(L-396)40.205.40 40 IF(L-433)45,210.45 45 IF(L—434)50.211.50 50 IF(L-435)55.212.55 55 IF(L-262)60.213.60 60 IF(L-590)65.216.65 65 IF(L-1239)70.220.70 70 IF(L-1201)75.294.75 75 IF(L-1202)80,296.80 80 IF(L-152l)85.297.85 85 IF(L-I522)90.298.90 90 IF(L-1523)95.299.95 95 IF(L-1524)100.300.100 100 IF(L-1329)105.301.105 105 IF(L-1330)110,302.110 110 IF(L-1331)115.303.115 115 IF(L-1332)120,304.120 120 WRITE(1,125) 125 FORMAT(3X’ILLEGAL COMMAND! TRY'AGAIN.'/) C GO TO 2 C HA 8 HALT C 170 STOP C C LP 8 LIST PARAMETERS C 189 WRITE(1.190)RHOI.RHO2.NT.THI.W2.TH2.W3.TH3.W§.TH§ 190 FORMAT(/5X.6HRHO 8 .F5.3/5X.6HRHO'8 .F5.3//5X.'OMEGA1 8 ' 1F5.3.10X'THETA1 8 'F5.3/5X’OMEGA2 8 ’F5.3.IOX‘THETA2 8 ' 2F5.3/5X’OMEGA3 8 ’F5.3.10X’THETA3 8 ’F5.3/5X7OMEGA4 8 ’ 3F5.3.10X7THETA§ 8 'F5.3/) C GO TO 2 C LF 8 LIST’FILES C 194 WRITEC1.195)FILE1.NOPTS.FILE2.IXMIN.FILE3.LTYPE.INTVAL 195 FORMAT(/5X7FILE1 8 'A6,10X’FILE LENGTH 8 ’I3/ 15X'FILE2 8 ’A6.10X’STARTING WAVELENGTH 8 'I3/ 25X'FILE3 8 ’A6.10XFFILE TYPE 8 'A2/ 195 329X'WAVELENGTH INTERVAL 8 ’I2/) C GO TO 2 3 GO 8 EXECUTE CURRENT INSTRUCTION SE'I' 200 IF(D202.202.201 201 GO TO (500.600.600.600).K 202 WRITE( 1 .203) 203 FORMAT(3X.'WHICH CORRECTION mDE. ST. EX. EM. OR SY7’/) GO TO 2 C C FL 8 FILE IMGTH C 205 READ (1.206)NOPT‘S 206 FORMAT( 13) GO TO 2 C C F1 8 DATA FILE FOR POSITION 1 c ’2 g I I I I 2 6 F3 3 I I I I 3 C 210 READ(1.215)FILE1 GO TO 2 211 READ(1.215)FILE2 GO TO 2 212 READ(1.215)FILE3 GO TO 2 C C DE 8 SET DEFAULT PARAMETERS C 213 RH01=.04 RHO28.04 W18.05 W2=.265 “3.70 W4=.915 TH18.05 TH2=.29 “3:07 TH48.94 NOPT‘S81 K80 LT'YPE81236 IXMIN8200 INTVAL=2 GO TO 2 215 FORMAT( A6) C C IN 8 WAVELENGTH INTERVAL C 216 READ( 1.217) INTVAL 217 FORMAT( 12) GO TO 2 C C SW 8 STARTING WAVELENGTH (EMISSION) C 220 READ( 1 .221) IXMIN 221 FORMAT( 13) GO TO 2 C C EX 8 EXCITATION PHASE C EM 8 EMISSION PHASE C SY 8 SYNCHRONOUS PHASE C 235 K82 LTYPE=L GO TO 2 295 296 297 298 299 300 301 302 303 304 500 600 196 ST 8 STATIC PHASE CORRECTION PARAMETERS READ(1.295)RH01 FORMAT(F5.3) GO TO 2 READ(1.295)RHO2 GO TO 2 READ(1.295)W1 GO TO 2 READ(1.295)W2 GO TO 2 READ(1.295)W3 GO TO 2 READ(1.295)W4 GO TO 2 READ(1.295)TH1 GO TO 2 READ(1.295)TH2 GO TO 2 READ(1.295)TH3 GO TO 2 READ(1.295)TH4 GO TO 2 CALL CHAIN('STAT') CALL CHAIN(’SPEC’) END 197 CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C C SUBROUTINE FOR KEYBOARD INPUT C C FOR PROGRAM FLUORO.FT C C C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC S OPDEF EEC 6032 S OPDEF BSW 7002 SUBROUTINE IN( L) S CLA S TAD K212 S JMS OUT S TAD K215 S JMS OUT S TAD K276 S JMS OUT S TAD K207 S JMS OUT S KCC S JMS GET S BSW S DCA TEMP S JMS GET S TAD TEMP S DCA I 1 RETURN SK276. 276 SK212. 212 SK215. 215 SK207. 207 SK77. 77 STEMP. 0 SGET. 0 SA. ESP 8 JMP A S KRB S TLS SB. TSF S JMP B 8 AND K77 S JMP I GET SOUT. 0 S TLS SC, TSP S JMP C S CLA S JMP I OUT STOP 198 gCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C PROGRAM STAT.FT C C STATIC ABSORPTION CORRECTION C C C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC. DIMENSION F1(200),F2(200).F3(200) COMMON RHOI.RH02,W1.N2.W3.N§,TR1.TR2.TR3.TH4 COMMON IXMIN,NOPTS,INTVAL.K;FILE1,FILE2.FILE3 500 CALL IOPEN( 'FLP2' .FILEI) DO 510 I=1.NOPTS READ(4.505)FI(I).F2(I),F3(I) 505 FORMATt3El5.9) 510 CONTINUE M80 515 Sl=0. 88180. 82:0. 882:0. 83:0. 88380. no 520 I=I.NOPTS SI=SI+F1(I) SSl=SSl+Fl(I)*Fl(I) S2=Sz+F2(I) SSZ=SSZ+F2(I)*F2(I) 83=83+F3(I) ssa=ss3+r3(l)*F3(l) 520 CONTINUE XN= FLOA’N NOP‘IS) SD1=SD(SSI,SI,XN) SD2=SD(SS2.82.XN) 803=SD(883.83,XN) Sl=Sl/XN 82:82/XN 83:83/XN IF(M)550.521.550 521 READ(1.522)OFILE 522 FORMAT(’OUTPUT FILE:'A6) CALL OOPEN(’FLP2’.OFILE) WRITE(4.530)FILEI WRITE(4,535)SI.SD1,82.SD2.S3,803 530 FORMAT('FILE: ‘A6/16X'RAV MEANS'I) 535 FORMAT(’F18 'E15.9' +/- 'E15.9/'F28 'E15.9' +/- 1'F38 ’E15.9' +/- 'E15.9/) M81 DO 540 I81.NOPTS CALL CORR(F1(I).F2(I).F3(I)) 540 CONTINUE GO TO 515 550 WRITE(4.555) 555 FORMAT(16X7CORRECTED MEANS’l) WRITE(4.560)Sl.SD1.S2.SD2.S3.SD3 560 FORMAT(’F8 ’E15.9' +/- ’E15.9/'A8 'E15.9' +/- ’E15.9/ 14HA’8 .E15.9' +/- ’E15.9/) CALL OCLOSE CALL CHAIN(’FLUORO’) END 199 CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC C C C C C CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC 600 615 640 641 642 650 700 701 750 755 PROGRAM SPEC.FT C SPECTRAL ABSORPTION CORRECTION C C DIMENSION F1(200).F2(200).F3(200).PM(501) COMMON RHOI.RHO2.W1.W2,W3.WI.TH1.TH2.TH3.THM COMMON IXMIN,NOPTS.INTVAL.K.FILE1.FILE2.FILE3 CALL IOPEN(’FLP2'.FILE1) READ(4.615)(F1(I).I81.NOPTS) FORMAT<17X.E15.5) CALL IOPEN('FLP2'.FILE2) READ(4,615)(F2(I).I81.NOPTS) CALL IOPEN('FLP2'.FILE3) READ(4.615)(F3(I).I=1.NOPTS) DO 640 I81.NOPES CALL CORR(F1(I).F2(I).F3(I)) CONTINUE IF(KF2)641.700.641 NSKP=IXMIN-199 CALL IOPEN(’SYS'.’PM’) READ(4.642)(PM(I),I=1.501) FORMAT(9(F6.3.2X)F6.3) D0 650 I=l.NOPTS Fl(I)=Fl(I)*PM(NSKP) NSKP8NSKP+INTVAL CONTINUE READ(1.701)OFILE FORMAT('OUTPUT FILEz'AO) CALL OOPEN('FLP2’.OFILE) D0 750 I=l.NOPTS XI8FLOAT( I) WRITE(4.755)XI.F1(I).F2(I).F3(I) CONTINUE FORMAT('RD'4E15.5) CALL OCLOSE CALL CHAIN('FLUORO‘) END 200 gCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCch g ABSORPTION CORRECTION SUBROUTINE g CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC SUBROUTINE CORR( Fl .Il'2.I"3) COMMON R1101,Rn02.WI.N2.N3.N4.TBI.Tfl2.m.TN4 COMMON IXMIN. NOPTS. INTVAL.K.FILEI .nmmum CLOG= ALOG< 10.) I’m: 1 ./( wa-wn SEXP= 1 . 41113-1111) T18(F1/F2) txPEXP 12H F3/E2) "SEX? PRCI=I.+RROI*T13*(2.-N3-N4) PR02=1. +RBOI*T1**(2.-NI-W2) SRCI= l . +RH02:T2**( 2 . -TII3-TII4) SRC2= 1. +RBOZ:T2**( 2 . -TRl-TH2) T1=TI*(PR02/PRC1)*:¢PEXP T2: T22“ SRC2/SRC I ) mkSEXP AP: -ALOG( TI ) ICLOG ASI-ALoct 12) /CLOG A: APazASM wz-m ) *( 1112-1111 ) 8CIOG=IICLOG B=T1**W2-Tlxa:w1 C- T2**TH2-'I2**THI FI=F1*A/(C*(T1**W4-TI*3W3)) F28 I-‘2tA/( C233) F38 F3*A/( BM mxxm4-T2st) ) RC=( I. -RROI*RHOI*TMTI)8( I .-RBO2*R302#T2*T2)/( I .-nnon/( I .-R1102) rlsrlxncxrnclxsncz F2=F2xRC/PRC2/SR02 F38 E3*RC/PR62/SRCI Fl=(F1+F2+F3)/3. IP2=AP F3=As RETURN END 2 E g 3 SUBROUTINE SD COMPUTE STANDARD DEVIATION FROM SUM AND SUM OF SQUARES CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC FUNCTION SD( SS .8. X) SD8 SQRT( ABS( SS-SaRS/X) /( X- 1 . ) ) RETURN END 000000 OOGOOO O