Jllfl'llfl'flmlililflllllfillllfifllfllfll 3 1293 00642 5767 / " LIBRARY , MichiganSmcc V"l)nW”€3§37 n.» - u' X This is to certify that the thesis entitled THE DEVELOPMENT OF AN ARRAY DETECTOR SPECTROPHOTOMETER presented by PETER JOSEPH AIELLO has been accepted towards fulfillment of the requirements for M. S , _degree in Mil-1112!... Major professor meo 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MTERIALS: ,,,, Place in book return to remove « charge from c1 rculat‘lon records , .- ¥- \ 4“"!!! THE DEVELOPMENT OF AN ARRAY DETECTOR SPECTROPHOTOMETER Bu Peter Joseph Aiello A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1980 / /,./ ABSTRACT THE DEVELOPMENT OF AN ARRAY DETECTOR SPECTROPHOTOMETER 139 Peter Joseph Aiello A microcomputer controlled linear diode array spectrophotometer has been developed. The linear diode array used has 512 individual light sensitive diodes each of which is sensitive to electromagnetic radiation with wavelengths from 200-1000 nm. A polychromator was designed using a concave holographic grating which helps reduce stray light levels and produces a flat Field image (400 nm) across the light sensitive area of the detector. The detector can be 'moved in the focal plane of the grating with no degradation of resolution. The inlet system optics were designed for maximum light intensity at all wavelengths. The linear diode array (LDA) is self-scanned to provide real-time. computer-compatible. serial electronic output For all 512 channels in succession. The microprocessor. an Intel 8085A: is used to collect and store the data on a floppy disk. The microcomputer also controls the detector integration time. By increasing the integration time. weak Peter Joseph Aiello signals can be enhanced but there is a loss of dynamic range as the dark current can approach saturation at room temperature in a time as short as 2 sec. Charge integration can be used to optimize the output of the detector while avoiding saturation. Resolution remains constant at all integration times as there is no evidence at interchannel cross-talk (blooming). The instrument has been applied to molecular absorbance spectroscopy. To Patsy ii ACKNOWLEDGMENTS My sincere appreciation goes to Professor Chris Enke for the help and encouragement that he has given me during the course of my research here at Michigan State University. I would also like to thank Professor Stan Crouch for many invaluable discussions. I would like to thank Michigan State University. Abbott Laboratories and the Office of Naval Research for the financial support they provided. I would like to thank the members of Professor Enke’s research group for their help and friendship. especially Hugh Gregg and Bruce Newcome. Finally. I offer my thanks and appreciation to my wife. Pat. for her continuous support and love. iii Chapter LIST OF TABLES . LIST OF FIGURES. CHAPTER I. INTRODUCTION . CHAPTER II. A FUNCTIONAL DESCRIPTION OF THE INSTRUMENT. A. Introduction . B. A Block Diagram of the Instrument. C. Microcomputer System Software. D. Data Analysis Software . CHAPTER III. THE DESIGN OF THE SYSTEM ELECTRONICS . A. Introduction . B. Linear Diode Array. Clock Drive and Amplifier Circuits . C. Microcomputer Interface to the RC-1024SA Amplifier Board. D. The Microprocessor and its Peripherals . CHAPTER IV. THE DESIGN OF THE SYSTEM OPTICS . A. Introduction . B. The Inlet System Optics. C. The Polychromator. CHAPTER V. CHARACTERIZATION OF THE LINEAR DIODE TABLE OF CONTENTS ARRAY SPECTROPHOTOMETER . A. Introduction . iv .vi vii .18 .18 .18 .24 .27 .30 .30 .30 .35 .35 Page Chapter B. Dark Current . C. Integration Time . D. Stray Light. E. Resolution . F. Saturation Effects . G. Diode to Diode Sensitivity Variations. H. Dynamic Range. CHAPTER VI. OPERATION AND APPLICATION OF THE INSTRUMENT. A. Routine Operation of the Instrument. .8. Application of the Instrument. APPENDIX A: SELECTED PROGRAM LISTINGS. A. Microcomputer Software . B. Minicomputer Software. APPENDIX B: INTERFACE SCHEMATIC DIAGRAMS . APPENDIX C: PERIPHERAL SCHEMATIC DIAGRAMS. APPENDIX D: GRATING SPECIFICATIONS . REFERENCES . .35 .40 .41 .47 .51 .55 .56 .59 .58 .61 .66 .66 .76 .83 .88 .91 .92 Page LIST OF TABLES Table Page 2.1 Current SLOPS library entries . . . . . . . .11 2.2 Current library entries for instrument operation . . . . . . . . . .14 vi Figure LIST OF FIGURES A block diagram of the LDA spectrophotometer. Absorbance spectrum of KMnO4. Pin configuration of the LDA integrated circuit. Sensor geometry and aperature response function . Equivalent circuit. Block diagram of interface between amplifier board and SDK~85 . Block diagram of the CPU and its peripherals. Photograph of the array and optical system. Approximate drawing of the polychromator. Spectrum of a neon filled hollow cathode with electronic background Spectrum of a neon filled hollow cathode with background subtracted out . Background signal vs. integration time. Detector output vs. integration time. Percent stray light . 0.1. lamp spectra at varying integration times 0.1. lamp spectra with 500 nm cutoff filter . vii Page 20 21 22 24 28 31 34 36 38 39 Figure Page 5.8 Spectrum of a neon filled hollow cathode with a 100 'pm slit width . . . . . . . . 48 5.9 Spectrum of a neon filled hollow cathode with a 50 um slit width. . . . . . . . . 49 5.10 Spectrum of a neon filled hollow cathode with a 25 um slit width. . . . . . . . . 50 5.11 Spectrum of a neon filled hollow cathode at an integration time of 100 msec. . . . 52 5.12 Spectrum of a neon filled hollow cathode at an integration time of 500 msec. . . . 53 5.13 Spectrum of a neon filled hollow cathode at an integration time of 1.0 sec . . . . 54 5.14 9.1. lamp spectrum showing sensitivity variations from diode to diode. . . . . . 57 5.15 Peak intensity vs. Ztransmittance. . . . . . . 58 6.1 KZCrzO7 working curve. . . . . . . . . . . . . 63 6.2 KMnO working curve. . . . . . . . . . . . . . 64 8.1 Interface RAM schematic diagram. . . . . . . . 83 B.2 Counters and decoding for interface RAM. . . . 84 B.3 Interface control schematic diagram. . . . . . 85 B.4 ADC and Integration control schematic diagrams 86 3.5 Control circuitry timing diagram . . . . . . . 86 viii Figure 4 k RAM board schematic diagram. 1 k PROM board schematic diagram . USART board schematic diagram. Grating specifications . ix Page 88 89 90 91 CHAPTER I INTRODUCTION Traditionally. the technique for the measurement of electromagnetic radiation in the UV-Vis region involves a dispersive element such as a grating or prism and a detector system such as a photomultiplier tube or photographic plate. Spectroscopists have long searched for an electronic measurement system that combines the simultaneous multiple wavelength coverage of the photographic plate with the real time electronic readout. large dynamic range. linear. response. and sensitivity of the photomultiplier tube (PMT)(1). The first and most common multichannel detector is the photographic plate. It has several advantages; a relatively easy channel identification. simple operation. minimal investment. integrating capability. and, it can be constructed in any size for any number of channels. Unfortunately. it has a limited dynamic range. a non-linear response and a time-consuming data retrieval procedure. One type of multichannel system is based on the use of a PMT array which is optically arranged across the focal plane of a polychromator (2). Advantages included wide dynamic range. excellent sensitivity and rapid response. However. the number of possible channels is limited. and the PMT’s must be arranged at wavelengths apprOpriate for predetermined specific applications. New detector systems based on modern electronic image sensors can provide over 1000 independent optical channels. These detectors. when interfaced to a computer. provide excellent flexibility in data handling. signal processing. modes of integration. data accumulation and real time display. Nevertheless. they have limited resolution (due to the number of discrete light sensing elements). lower sensitivity than the PMT. incomplete readout (lag) and interchannel crossvtalk (blooming). The great potential of multichannel array. or TV type detectors such as charge coupled devices (CCD). vidicons (V) and linear diode arrays (LDA) for spectrometric measurements has been discussed in numerous recent reviews (3-10). In atomic spectroscopy. these detectors have been exploited primarily to carry out multielement analysis (9-16). For atomic absorption spectrometry. where light levels are reasonably high. much success has been achieved (15-18). For atomic emission or flourescence spectrometry. where light levels are much lower. Ninefordner and co-workers (11.19) have concluded that multichannel detectors will find more limited use because of seriously degraded detection limits. However. in many cases where analyte concentrations are well above conventional detection limits. multi~element quantitative analysis with multichannel detectors by atomic emission has been demonstrated to be quite feasible (13.14.20.21). In atomic work. multichannel detectors have also abeen employed for profiling (22.23). correlation (24). time-resolved studies (25).spectral stripping (26) and internal standard compensation (27). In molecular spectrometry. these detectors are not used as much for multicomponent analysis because molecular bands are much broader than atomic lines. However. Pardue. Milano and co-workers (28.29) have shown that multicomponent absorbance analysis is feasible for a few components in certain cases. In molecular work. these detectors are often employed as one means to construct a rapid scan spectrometer to monitor the output from an HPLC or GC (11.30-33). One type of electronic image sensor is the self-scanning linear array of silicon photodiodes. This dissertation describes the design and characterization of a microcomputer controlled linear diode array detector spectrophotometer. The instrument was designed for multi-component determinations with absorbance values from O to 2.0 or more. Many of the properties of the polychromator and the detector are measured as these properties ultimately limit the performance of the instrument. Chapter II presents a general overview of the instrument. Several important features of the linear diode array (LDA) spectrophotometer including the inlet system optics. polychromator and detector are described. Also the microcomputer system and the software system are presented. Chapter III discusses the design of the system electronics. Included here is a discussion on the control circuitry and operation of the LDA. the interface between the LDA amplifier circuitry and the microprocessor. and other peripherals also connected to the microprocessor bus. Chapter IV presents the design of the system optics. The design of the inlet system optics was chosen for maximum light intensity over the entire wavelength region from 300-1000 nm. The polychromator was designed for a minimum of stray light and a flat field image across the light sensitive elements of the detector. A maJor goal of the research was the characterization of the LDA spectrophotometer. The experiments and results are discussed in Chapter V. Properties of the polychromator examined include spectral region covered. resolution and stray light. Characteristics of the LDA examined are integration time. dark current. and saturation effects. Chapter IV explains the routine operation of the instrument and its application in molecular absorbance studies. CHAPTER II A FUNCTIONAL DESCRIPTION OF THE INSTRUMENT A. Introduction In this chapter a general overview of the instrument is presented. First. each basic part of the instrument is -described in the context of its relationship with the remainder of the instrument. Second. the microcomputer software and its use are discussed. and further software which was used on the PDP 11/40 minicomputer for data analysis and plotting of the acquired data is also described. B. A Block Diagram of the Instrument The spectrophotometer discussed here is different from a traditional spectrophotometer in one most important aspect. This is the fact that light intensities over a large spectral region can be monitored in the time a normal spectrophotometer takes to produce results for a single wavelength. This is achieved through use of a linear diode array (LDA) which was purchased from Reticon Corp (34). A block diagram of the instrument can be seen in Figure 2.1. The linear diode array. which is the distinguishing feature of this spectrophotometer. is a Reticon type RL-512S. This was the first solid state image sensor designed specifically for spectroscOpy applications. Some SOURCE // / CELL FOCUSING LENS ENTRANCE SLIT REFLECTING MIRROR HOLOGRAPHIC GRATING UNEAR DIODE ARRAY L DRIVE + , AMPLIFIER RAM + cnecun PROM I MICRO K N | ‘ INTERFACE—W__USART mm“ L—F—J __..__J 1 .___J FLOP-PT DISK .L..—__ Fig. 2.1 A block diagram of the LDA spectrophotometer important features of this device are: 1) There are 312 individual sensor elements (each 25 microns wide by 2.5 mm high) which allow the spectrum to be divided into 512 simultaneously exposed channels. 2) The LDA is self-scanned to. provide real time computer compatible serial electronic output for all 512 channels in succession. 3) The variable integration time (from 2.3-400 msec at 20°C) can provide a dynamic range greater than l0.000:1. The integration time can be optimized depending on the wavelength and the light level. 4) The detector is responsive over a range of 200-1000 nm. When using a device such as this linear diode array which can generate data at extremely fast rates. it becomes apparent that the data acquisition is best done by computer. For the operation and control of this instrument. an Intel 8085A microprocessor was chosen. It is an 8-bit microprocessor that is powerful enough to fully control the instrument functions as well as perform the data acquisition process. The microprocessor and a variety of interfacing integrated circuits have been designed to make interfacing relatively simple. Two kilobytes of programmable read only memory (PROM). 4 k-bytes of random access memory (RAM) and a universal synchronous/asynchronous reciever transmitter (USART). all designed by Bruce Newcome (35). have been built and implemented in this system. Operator interaction and control of the instrument is provided through a Hazeltine 1400 terminal which is connected through the USART to the microprocessor bus. Also interfaced to the bus is a Persci dual floppy disk system. This magnetic medium provides storage space mainly for data files but also for software. The floppy disk is currently the only link between the spectrophotometer and our research group’s PDP 11/40 minicomputer where further data analysis is performed. The microprocessor is also directly interfaced to the linear diode array. Data are acquired by first converting the analog video signal to a sequential digital output which can be stored in RAM. The conversion time for the 12 bit DATEL EH12B analog to digital converter is less than 4 microseconds. The integration time is the time between the readout of a specific diode from one scan to the next. This integration time can also be. selected by the operator through a command to the computer. Every spectrophotometer needs not only a good means of data acquisition and subsequent analysis. but also a very good optical design. The maJor optical components of the polychromator in this spectrophotometer are a concave holographic grating. a reflecting mirror and an entrance slit assembly. All of these were purchased from Instruments S.A. Inc. The grating provides a flat field image with minimum stray light. The grating was specially made (F 2.0) for the desired dispersion (400 nm) across the 512 elements of the linear diode array. The source optics were designed for high light intensity so that the absorption of highly absorbing materials could be measured. A General Electric #1974 lamp was chosen because it has a very compact filament and sufficient output in the visible region of the spectrum. The light then passes through the cell (in its constant temperature incubator) and is then focused onto the entrance slit with a double convex lens. The order of these Optical components is different from that in spectrometers with single-channel detectors. In any array detector spectrometer the dispersive element must directly precede the detector. This requires the cell to precede the entrance slit in the optical path. The fused silica lens (F 1.5) is closely matched to that of the grating to optimize light throughput. C. Microcomputer System Software A microcomputer software system could be either a single large program or at the other extreme. a collection of subroutines. However. all software systems must allow ease of instrument operation and data storage. In cases where many different types of experiments will be conducted. it is advantageous to be able to reprogram quickly. This is 10 possible in a subroutine structured software system. Listings of all the routines mentioned in this section can be found in Appendix A. The software system is called Structured Library Oriented Programming System (SLOPS). This operating system. as well as all the general use commands listed in Table 2.1. were written by Hugh Gregg (36). This software system allows the user to "create" new entries in the library of commands. Each entry can do a specific task or can call any or all of the other library entries as subroutines. For example. when you type the DISPLY command on the terminal. the microprocessor first looks through its library and if it finds DISPLY it begins execution. Upon completion of the DISPLY routine the computer will prompt SLOPS) and wait for a new input command. All library entries are assigned a specific address in RAM when the routines are compiled. It is through this "stack“ that the cpu searches its library entries. If one subroutine calls another. the compiler gives the address of the called subroutine. It is then executed and returns to the original subroutine. A number of library entries were created specifically for use with the linear diode array spectrophotometer and these are listed in Table 2.2. These subroutines are easily incorporated as library entries in SLOPS and are easy to use. The SCAN routine enables the interface to collect and 11 TABLE 2.1 Current SLOPS Commands Command Subroutines called Action ADDR Returns address of library entry TOASCI Binary to ASCII UNASCI ASCII to Binary BLANK TTYOUT Blanks terminal screen BRKDNN DCMP Breaks input line into words CHARIN Inputs single character CHECK Compares current word in library CLRDSK Disk data sink CONVERT NUMBER.CVTEXT Converts any number to PUSH2.PRINT any base PRINTR.CRLF CRLF TTYOUT Sends CR+LF to terminal CVTEXT PUSH1.DIV Converts number to ASCII TOASCI.DCMP string. any base CVTINT DMULT.UNASCI Converts ASCII string to number DCMP Compares (D.E) to (H.L) DDIV DSUB (B.C)=(H.L.B.C)/(D.E) Command DISK DDATUM DDATA DCMD DNAIT DINCHR DIV DMULT DSUB INIT KERNAL LINK MULT NUMBER PRINT PRINTR SEARCH 12 Subroutines called PUSH2.PRINT TTYIN.CHARIN DINCHR.DDATA DDATUM.DCMD.DNAIT DNAIT DDATUM DHAIT DDIV MULT BLANK.KERNAL PUSH2.PRINT TTYIN.BRKDNN NORD.SEARCH.ADDR CHECK.LINK Action Disk monitor routine Outputs data character to disk Outputs data string to disk Outputs command to disk Waits till disk ready Gets character from disk (D.E)=(D.E)/(A) (D.E)=(A)*(D.E) (H.L)=(H.L)-(D.E) Initializes system Main monitor Library pointer update (D.E)=(D)e(E) Gets number on stack Prints message on terminal from top of stack Prints ASCII characters found on top of stack Searches library for command Command POPI PUSHl POP2 PUSH2 TTYIN TTYOUT WORD 13 Subroutines called Action Pop 1 byte off stack Push 1 byte on stack Pop 2 bytes off stack Push 2 bytes on stack Accept and echo input line Send character to terminal Pops pointer to next word off stack 14 TABLE 2.2 Current Library Entries for Instrument Operation Command SCAN DISPLY STORE CDATA SIGAVG INT PLOT Subroutines called CVTEXT.PRINTR DCMP.TTYOUT.CRLF CHARIN.PUSH2.PRINT CLRDSK.PUSH2 PRINT.DDATUM TTYIN.DDATA.DCMP DINCHR.TTYOUT PUSH2.PRINT.SCAN DISPLY.PRINT2 TTYIN.STORE DCMP.PUSH2 PRINT.TTYIN BRKDNN.NUMBER DISPLY.STORE PUSH2.PRINT.TTYIN BRKDNN.NUMBER DMULT.DIV DCMP.DSUB TTYOUT.CRLF Action Enables interface and collects data Display data on terminal Stores the data on disk Combines the SCAN. DISPLY. and STORE routines Signal averages user entered number of scans Sets the integration time Plots the data on the terminal 15 store the data from the linear diode array in RAM. The DISPLY routine displays the data in base 10 on the terminal screen in a 51 row by 10 column format. The STORE routine copies the data from RAM to a file on the floppy disk after asking the user for a file name. The routine CDATA basically calls all the above routines for easier operation. SIGAVG. a signal averaging routine. asks the operator for the number of scans to average. then collects the data. averages and displays the averaged data on the terminal. Another routine. INT. asks the user for the desired integration time and sets it accordingly. A final routine. PLOT. plots the data in a 24 by 64 character grid on the terminal screen. D. Data Analysis Software. As mentioned before. all data analysis software consists of routines created on our PDP 11/40 minicomputer. The data files stored on a floppy disk are first transferred to the minicomputer. They are currently in an unformatted binary form. After loading a data floppy on the 11/40. the routine ABSORB can be executed. It calculates the absorbance at each wavelength according to the following equation: AI = -logESI-DI]/ERI-DIJ where 51 is the signal intensity at wavelength I. R1 is the reference intensity with a reference solution in the sample 16 cell..and DI is the signal obtained when no light is falling on the detector (dark current). The absorbance file is created and stored. Before plotting. the file must be converted from an unformatted binary file to an ASCII file. The routine CONVERT is used for this conversion. At this point. readily available plotting routines can be used. One of these is MULPLT. written by Dr. T.V.Atkinson. which can plot intensity or absorbance vs. wavelength on either the terminal or the line printer. Figure 2.2 shows a typical absorbance spectrum of KMnO4 . ABSORBANC E 17 0.9 -— O.8 - (16 * 0.5 -* 0.3 - 0.2 - 0-0 I I l I I I 400 450 500 550 600 650 700 WAVELENGTH (nm) Fig. 2.2 Absorbance spectrum of KMnOy CHAPTER III THE DESIGN OF THE SYSTEM ELECTRONICS A. Introduction With an instrument such as this linear diode array spectrophotometer. data can be produced at an incredible rate. To avoid loss of information. it is necessary to collect and store these data at the same rate they are produced. This requires the use of a computer. Microcomputers today are very cost-effective. The instrument can be automated with one of these processors very easily. This might allow both the operator and larger computers to spend their time on bigger and better things. B. Linear Diode Array. Clock Drive and Amplifier Circuits The RL512S array as well as the clock drive and amplifier circuits were purchased from Reticon Corp (34). The Reticon RL512S is a monolithic self-scanning linear photodiode array. This device consists of a row of silicon photodiodes. each with an associated storage capacitor on which the photocurrent is integrated. The multiplex switches at each photodiode are switched in sequence by an integrated shift register scanning circuit to provide a serial output of the charge accumulated at each diode. The array is packaged in a 22 lead dual in line integrated 18 19 circuit package with a polished quartz window face plate. The pin configuration is shown in Figure 3.1. The diode elements are on 25 micron centers which correspond to a density of 40 diodes/mm. The overall length of our 512 diode array is thus 12.8 mm. The height of each diode is 2.5 mm. which gives each element a slit-like geometry with a 100:1 aspect ratio. This is desirable for use with a spectrograph. The sensor geometry is shown in Figure 3.2. The diodes consist of diffused p-type bars in an n-type silicon substrate. Light incident on the sensing area generates charge which is collected and stored on the p-type bars during the integration period. The accumulated charges are then sequentially switched into the video output for readout. The n-type as well as the p~type silicon surface is photosensitive. Light incident on one of the p-regions will generate charge which is stored on that diode. Charge generated by light incident on the n-type surface between 2 p-regions will divide between the adJacent diodes to produce the response functions shown in Figure 3.2. A simplified equivalent circuit of the linear -diode array is shown in Figure 3.3. Each cell consists of a photodiode and a dummy diode. both with associated storage capacitance. These diodes are connected through MOS multiplex switches to video and dummy recharge lines. One pair of recharge lines is common to all the odd elements and another pair is common to all the even elements. 6N0 SUBSTRATE EVEN START EVEN g5, EVEN 4:, SUBSTRATE EVEN EOL EVEN RESET GATE EVEN DUMMY VIDEO EVEN ACTIVE VIDEO RESET BIAS Fig. 3.1 @QNGUOUN— 5 Nc 6N0 ooo START 000 ch ODD 95, SUBSTRATE ODD END OF SCAN ODD RESET GATE ODD DUMMY VIDEO 000 ACTIVE VIDEO SUBSTRATE Pin configuration of the LDA integrated circuit 21 NH fi§§/% 2.5urn «Bum-H *—— 25prn -—>* . 1‘ _ ( SILICON DIOXIDE 3pm I’ I (R-SILICON) L ['5‘ I N-SILICON j 0'4“” N-I DIODE N MI 6 o ~e RESPONSE (l o 32 POSITION OP INCIDENT LIGHT SPOT Fig. 3.2 Sensor geometry and aperature response function awauefiu ucwwm>flovm W.m ONO.) >ZZDO ZN>MA are wkdw Pmmmm zw>m Owe.) m>_._.U< zm>wh - M... weqmemmamxl msa ewmwm II. 1.... N mzfi mo ozwil— awkmamm FEIm Zw>m ,1 e ; 1.5.4.5 H.) i .1.— .1 O mwooa oeoxe ._. I: >228 . I - . . - HG .fi : r . H smog“. . e m mwoflm>meoxe . - - .. - I .mwooa 9.0an H mac. n _ ooo 1. ._ 1 Que.) w>_._.U< COCA I .——-I > k) «b wkdw. kmmmm 000 u I «L an i mwogo ObOIn. 2:230 OmQ> 22230 000A I [HI m2] “.0 02w Al 4L er- mwkmawm ._.n:Im 000 inlA hmdhm III we 23 The multiplex switches are turned on and Off in sequence by shift register scanning circuits which periodically recharge each cell to 5 volts and store a charge 0(sat). on its capacitor. The shift registers are driven by multiphase clocks with periodic start pulses introduced to begin each scan. The cell-to-cell sampling rate is determined by the clock frequency. The integration time is the interval between start pulses. During this integration time. the charge on each capacitor is gradually removed by the reverse current flowing in the associated diode. The reverse current is made up Of two components. the photo current ip. and the dark leakage current id. The photocurrent is the product of the diode responsivity and the light intensity or irradiance. During a line scan time. the charge 0 removed from each cell is the product of the photocurrent and the line time. This charge must be replaced through the video line when the diode is sampled once each scan. Thus. the output signal obtained from each scan is a train of 512 charge pulses each proportional to the light intensity on the corresponding photodiode. By properly phasing the clock drives to the two shift registers all of the diodes can be sampled in proper sequence. The two video lines can then be simply connected together to provide a continuous train of output charge pulses. In addition to the signal charge. switching transients 24 are capacitively coupled into the video lines by the multiplex switches. These same transients are introduced into the dummy lines and therefore can be eliminated by reading out the video and dummy lines differentially. Output pulses are provided when the last Odd and even elements are sampled by the shift register scanning circuit. The RC-1024SA interface board has 5 outputs: 1) Two end of line outputs for even and odd shift registers 2) Start output to monitor the beginning of the scan 3) An oscillator clock output 4) A combined even-odd video output C. The Microcomputer Interface to the RC-1024SA Amplifier Board The interface discussed here was designed and built for use between the SDK-BS microcomputer and .the RC-1024SA amplifier board. A block diagram of the circuit is shown in Figure 3.4. All circuit diagrams can be found in Appendix B. The basic operation of this interface includes a direct memory access (DMA) circuit which loads the output of the analog to digital converter (ADC) directly into RAM after which the microprocessor can access the RAM for any other purpose. To initiate data collection. a chip select (CS) and high DO signals are sent to the control logic. At this 25 mmeom ecm semen ceaeaaeEm ceeaeme menaceuca Co emcemae eucam e.m wOOOwo wwmw‘flo mmwmooq 29.2532. 0.00.. JOKPZOU .5..de hug—h 004 z. @3424 Dmdom mm.......n.n:>.< don. point the microprocessor is put on hold and the control logic waits until a start pulse is found. This pulse comes from the Reticon amplifier board. Once this happens. the control logic sends out start conversion. address increment. write enable. chip select. and latch enable pulses. All these are timed properly to load each converted data point into its sequential place in RAM. See the timing diagram also in Appendix B. Immediately before the ADC however. is a simple amplifier with gain which converts the O to 3 volt output (no light to saturation) of the detector board to the 0 to 10 volt input range Of the converter. The conversion time of the Datel EH12B2 ADC is 4 n-sec. This allows a maximum theoretical clock rate of 1 MHz. The actual clock rate is 800 KHz. At the end of a complete scan. the even end of scan signal shuts off the control logic and releases the microprocessor to continue execution of any program. The integration time can also be controlled by the microprocessor. The array amplifier board has a series of 3 4-bit synchronous counters which delay the start pulse by the amount of time set at the preload switches. These preload switches were removed and a set of latches tied to the data bus were installed in their place. These allow the computer to vary the integration time from 2.7 msec to to 1320 msec in 5.2 msec increments. D. The Microprocessor and its Peripherals The 8085A microprocessor was chosen because of its high performance 1(1.3 u_sec instruction cycle). popular instruction set. and low cost. This processor is included in the MCS series System Design Kit (SDK—BS). The SDK-85 was used because it is a complete single board computer with memory and I/O. Included in the SDK-85. which was purchased from Intel Corporation. is the 8085A CPU. 2 kilobytes of Read Only Memory (ROM). 256 bytes of Random Access Memory (RAM). 38 paralell I/O ports. and 1 serial which utilizes the SID.SOD pins of the 8085A and which has a software generated baud rate of 110. The SKD-85 also has an interactive LED display and keyboard with extensive monitor software in the 2 k of ROM which acts as an excellent front panel (useful in debugging custom interfaces). Figure 3.5 is a block diagram of the SDK-85 and all the peripherals in the system with the exception of the previosly described interface to the LDA amplifier board. Appendix C contains the schematic diagrams for the peripheral boards which Bruce Newcome designed (35). There are 4 k bytes of RAM. 2 k bytes of Programmable ROM (PROM) and 1 Universal Synchronous/Asynchronous Reciever Transmitter (USART) connected to the bus in. the micro system. The PROM is Used for the storage of all the SLOPS routines listed in Table 2.1. The RAM is used for all user written routines. The USART is used for serial mHmezafiLwa my“ new 3&0 wzu e0 Ememmfiu zoofim m.m .mwu _ w>ED v.90— u. 32.2mm... _II|III.II. Ox. . 24m .2de .543 " mwtfi emu + 9.4099. n m _ H.— Q - _ mom 555 v— HH _ HH _ 23. whose some _ a .88 _ mwcrm x e. mwtrm x N _ sou _ _ mm I v.0 m _ 29 communication to the Hazeltine 1400 terminal. The floppy controller is a Persci Model 1070 which has its own 8080A processor and programs which control the connected Persci Model 760 dual floppy disk drive (37). CHAPTER IV THE DESIGN OF THE SYSTEM OPTICS A. Introduction In this chapter both the inlet system optics as well as the design and construction of the polychromator are discussed. The inlet system optics were designed for high light intensity from a near point source. This facilitates experiments where interest is focused on samples of high absorptivity as the detector sensitivity is not as great as a photomultiplier tube. especially in the ultra-violet region of the spectrum. The polychromator design centers around the use of a holographic grating. The grating is designed to produce a flat field image across the photosensitive elements of the detector with a minimum of stray light. Figure 4.1 is a photograph of the spectrOphotometer optics. The polychromator and inlet system optics are mounted on a pair of perpendicular optical rails. This arrangement provided the necessary flexibility during the development of the spectrophotometer. B. The Inlet System Optics The design of the inlet system Optics was selected so that the light intensity at all wavelengths generated by the source is maximized at the detector surface. The source used is a General Electric #1974 quartz iodide lamp. It is 30 31 r... __e v» . . . ‘w'n-“w‘w‘, i . Fig. 4 1 Photcgragw a? the array and optical system a 20 watt lamp which uses an input voltage of 6 volts. It was chosen because of its very compact filament ( 1mm by 2mm ). and because it is intended for normal spectrophotometric use. The light then enters the cell assembly. This assembly.taken from an Abbott Laboratories VP Clinical Analyzer. supports a doughnut shaped ring of individual sample compartments which allow for easy determinations of multiple samples. The sample cell assembly also contains a heating coil and thermistor for maintaining a constant temperature of the sample (35°C). As the light exits from the cell assembly. it disperses. To collect the dispersing light. a simple double convex lens is inserted into the light path. It is a fused silica lens with a 50mm diameter and a 75mm focal length (F 1.5). The lens focuses all the collected light onto the entrance slit. This lens also serves another purpose; the F number of the lens closely matches that of the grating. which allows full use of the grating surface (i.e. light intensity at the detector surface is maximized). C. The Polychromator A polychromator is a spectrum dispersing device which produces simultaneous spectral information over a bond of wavelengths. The entrance slit assembly. purchased from Jobin Yvon 33 Division of Instruments S. A. Inc.. has interchangeable slits of 25. 50. 100. and 1000 IIm widths. The model H2O slit holder with fishtail was modified to fit onto the optical rail. Mounted in the same assembly as the entrance slit is a front surface mirror which serves only to move the source optics away from the detector. An approximate drawing of the spectrograph is shown in Figure 4.2. In a linear diode array spectrophotometer such as this one. it is most important to have a grating that can produce a "clean" spectrum on a flat surface. The linear diode array is then positioned in this plane so that the optical quality of the image is identical from one end of the spectrum to the other. One of the newest advances in holographic grating design technology has made possible the production of flat field spectrographic. concave. aberration-corrected gratings which by themselves produce a flat spectrum. This spectrograph. in addition to permitting the use of the full advantages of multiwavelength detection. has two other advantages; its extremely low stray light level. and its extreme simplicity. The concave. holographic grating. also purchased from I. S. A. Inc . has a focal length of about 100mm and a diameter of 50mm (F 2.0). It is also mounted onto the optical rail. Specifications related to the grating can be found in Appendix D. The array is mounted in a similar manner in the focal plane. 34 FOCAL PLANE ENTRANCE APERTURE Fig. 4.2 Approximate drawing of the polychrovator CHAPTER V CHARACTERIZATION OF THE LINEAR DIODE ARRAY SPECTROPHOTOMETER A. Introduction In this chapter the characteristics of the linear photodiode array spectrophotometer are discussed. The polychromator properties that are emphasized here include spectral region covered. resolution and stray light. The linear diode array characteristics emphasized' are integration time. dark current and saturation effects. B. Dark Current One of the maJor operational characteristics of the linear photodiode array is its electronic background noise. This background signal can be seen in the spectrum Of a neon filled hollow cathode lamp shown in Figure 5.1. The slit width used is 25 um and the integration time is 200 msec. The electronic background. or dark current. is the maJor source of the overall height of the pedestal Observed in Figure 5.1. The stray light can also be seen. However. it is a small percentage of the total height. The noise across the array is due to diode to diode dark current variations. A complete dark current spectrum is Obtained by blocking the entrance slit. collecting the data and storing them on a floppy disk. At any given integration time. the dark current is very reproducible. At an integration time 35 INTENSITY (relative) 36 (,n (j C. J J I - I I _ I I J I II . I I -I II I I000 ~ I III I I I‘ II I ‘ II I III I _ II II II II II? . III II III! IIIIII - - II III III - II III III ' III IIIIIII I‘i‘ 1 I} IIIIII.'I I IhI“ IIIIIIVU‘ ' 1‘ WWII I“myII.‘.""“,-~{‘."‘ ‘I I.‘ . l ‘ II I h I ' I I - rt "MWMWNMMQA II NMWfimwmfiiva‘ 500 - I ' I I _ I ‘ I O lllliIllIIIllIIIIIIIIIIllIlllI’IlllIIllIlIllIIj 250 500 350 400 450 500 550 500 650 700 HPNELENCTH (nwfi Fig. 5.1 Spectrum of a neon filled hollow cathode with electronic background 37 a? 50 msec. the dark current in a given channel has a maximum relative standard deviation of 1.29%. The ZRSD is calculated at all wavelength channels from 5 collected scans and the reported value is the maximum observed at any channel. As the dark current is very repeatable From scan to scan. it can be removed From any spectrum by subtraction. This operation is easily accomplished by soFtware on the 11/40. The neon spectrum shown in Figure 5.1 is shown in Figure 5.2 minus the dark current. Even if the spectral signal is so small as to be essentially obscured by the background. subtraction can still recover a useful spectrum. Except where noted. all spectra shown in this thesis are background subtracted spectra. The presence of the dark current can severly limit the effective use 0? the integrating capability of the array. In an integrating type of detector. such as the linear photodiode array. the charge From the dark current increases as the integration time increases and can. at integration times longer than 2 seconds. completely saturate the array (i.e. completely discharge the reverse bias on the diode). This leaves no dynamic range for signal measurement. The background signal as a function of integration time of one photodiode is shown in Figure 5.3. At an integration time of 10 msec the array dark current is nearly zero. At an integration time of 600 msec. Just over one third of the INTENSITY (relative) 38 -——o—- ——.-—— I 700 600 j 3 .I 500—3 1 400-3 l 1 300 j I 200—} II 100-3 III I I I I I 3 III I : :19: I flIIII I III 0" IIIIITTIITIIII 250 300 350 400 450 500 550 600 650 WAVELENGTH (nm) Fig. 5 2 Spectrum 0? a neon Filled hollow cathode with background subtracted out DARK CURRENT (relative) ‘ 39 2000 ~ —I 1000 — - / .I 4 O T'lITIIITTIITTTTTTfTITITII7TTTIllTlIITTIITIIIIIIIII O ICC 200 300 400 500 600 700 800 900 1000 INTEGRATION TIME (msec) Fig. 5.3 Background signal vs. integration time 4O useful dynamic range is taken up. At an integration time of 1 sec only one third of the dynamic range is left for the signal and at an integration time of over 2 sec the array is completely saturated; hence no signal can be measured. The non-linear response is most probably due to a charge leakage from the video recharge line on the LDA integrated circuit. It is expected that cooling the array will reduce this charge leakage to a negligible value and the plot will approach linearity. The dark current level of the array can also be significantly reduced by cooling. According to the literature (6). if the array were cooled to about -15"C. the dark current at an integration time of 15 sec would be the same as that for an uncooled array at an integration time of 45 msec. C. Integration Time Charge integration is a signal enhancement technique which takes advantage of the integrating capability of the LDA. The integration time is the time difference between the readout of a specific diode from one scan to the next. Charge integration can be a very useful signal enhancement technique as the S/N (signal to noise ratio) under many conditions increases linearly with the integration time. while the S/N increases only with the square root of the number of scans averaged (38-41). As mentioned before. the integration time is under microcomputer control. Detector 41 response (charge integrated) should be proportional to the integration time. Figure 5.4 shows a plot of detector output at 650 nm of the quartz halogen lamp versus integration time. The integration time was varied from 7.85 msec to 983.5 msec. The slit width was 25 pm. The intensity of the line was reduced using a neutral density filter which allowed 1.0% of the light to pass. The detector output was recorded for 5 scans and the standard deviation (SD) calculated. The SD never exceeded 11.5. The non-linearity is again explained by the charge leakage in the LDA and is expected to improve with cooling. D. Stray Light One of the principal reasons for using the previously described holographic grating is its low stray light levels. To measure stray light. a monochromatic source at a known intensity is used. The intensity of radiation observed at the other wavelength channels is measured and the stray light is reported as a percent. Figure 5.5 shows the percent stray light at all wavelength channels which results from a HerNe laser as the source (632 nm). An integration time of 1.068 sec was used. The intensity of the laser line was calculated by measuring the intensity at a minimum integration time and extrapolating. This is possible as the detector response is related to the integration time as DETECTOR OUTPUT (relative) 3000— I .I 2000* .1 .I 1000— ' / .I O l—TT—I'IFTTTTITTTIIITTITTTIIITITTTTTTIITTIITIITTTTITI O 100 200 500 400 500 600 700 800 900 1000 INTEGRATION TIME (msec) Fig. 5.4 Detector output vs. integration time PERCENT STRAY UGHT 0.04 0.01 0.00 43 —-‘ I L L ] l L 1 I J L l I 'II I MIMI III, IIUM II/III IIIIIIWWMI’I Illlllel‘IIIIIIITTIIIIIIIIITIIIIIIIIITIIIIII 250 300 350 400 450 500 550 600 650 700 WAVELENGTH (nm) U! U1 Fig. Percent stray light 44 shown in the last section. The spikes seen in several of the channels of Figure 5.5 are due to reflections inside the polychromator. Even at strag light levels as low as those shown here. it can be seen that at long integration times. errors in measurements can easilq be observed. Because the polychromator is after the sample in the optical path. strag light is produced at all wavelengths that the source emits. For experiments in the ultra-violet region of the spectrum. strag light from the red region of the spectrum where the source intensity is much greater (in the case of the quartz halogen source) becomes a serious problem. In Figures 5.6 and 5.7. spectra of the quartz halogen lamp are shown without and with a red cutoff filter in place. Each plot shows multiple spectra at integration times of 5. 10. 20. 50. 200 and 500 msec. At 300 nm. the source intensitg is near 0. However. it is easilg seen that light is being detected (Fig 5.6). This light is stray light due to all wavelengths entering the polychromator. Even with the red cutoff filter in place (Fig. 5.7). the stray light can be seen (300 nm). but it is significantly reduced. This greatlg increases the dgnamic range which is of great importance for experiments in the ultra-violet part of the spectrum. The saturation level appears to decrease in the higher integration time spectra of Figures 5.6 and 5.7 because the dark current level. which is increasing. is 45 4000 1444 ,. . a .. u ' M‘"7';.' l . 1'3)?- ' , “grace-n- "' ' . ,fi‘If ' ' u 4' - . Q C . . , : a . ' _ - , “sac...- name-{w : .(‘a «4 ""' “'- ""‘-'"‘ . f . Ci" ‘ " I -~ ‘v‘ «4.91“ . v . , "J 'l‘v.- ' . o“‘¢‘.’ 'rs A«\"' ‘a” ' 0 v, .' 1 . O . ‘ - ‘9Y2.‘. . . ‘ ‘ o ‘ as ‘ I . ‘1' ' o q o a s' I . INTENSITY (relative) N o o o I O I / l' 9 q, - " I o ‘. v ‘5 O / " b .- , h .. f t. . a / “ '5‘ ‘ w . g. . 'g f “ ‘ .‘k‘s .. . .3. O—erfiw+‘#¥<§filr?fiél.lIll}!IIWWWIIITIIIIIITIIIIIIITIII 250 300 350 400 450 500 550 600 550 700 WAVELENGTH (nm) Fig. 5.6 0.1. lamp spectra at verging integration times INNENSHV'(nflafiV€) 14 #14 L 250 llllj 300 350 46 ....;r:e.-»~-»- “all"? .‘K' , .. vrM" v . . ,‘.; g. 0.- -- .4- \ .' 5'3'1‘5'3'.‘ F" ' ' u o . .“.’.'l‘\'tc ' ’." e . p \I ' c 1‘ . . 'n . ’v . 0“ {3... “1.33.3.0. W ’ \ . 1- . ' , , . ,v. , . e . » p"“'¢.~'un" ”1",! ‘ - . .- s‘o’. ‘e ...J_\.¥' 'fi’, - ' .‘r'r‘yi‘f- ” v"’”.~ _ X.r.'--“ I '- -\" ,"a's" . d ' I . - _ ': , 0 . ‘ I . C e o . u . ' I ' I D I ‘ . l.- .. -..-.W-».~~.--- w“... x , .' .~J~.' . . . . ' -~d.'v~.- . _.. . . . 5 . - ' p . .. . " l' - ~‘.-~‘\:J._‘z\..”" - - “I“ \.".J . a" ‘1' ‘- "331"“. {1...}. _ "_ ~ ‘. . v . ls' ' ' ’ . . . _ I .2". 5‘ ‘ . I ‘ . .i .‘ ' . O ’ . max, '. I .' ' . ‘I 0.. . ' O ' ' . ' n e ' a , ' . . C ’ . . 0 I ' e ' . e I. ‘- I ' ' o . :. . . . . . . ,° ' C . . . ' ,YA . ' ,' .\ . s - . e f . . v i' ' A ’ . q n . u . .I ‘Q" I ' I Q ‘ O I' ’- 0 U - e I “ I . C ' u :. f 1 . F . I I . ' I ' I. . . . I I 'I ' . ' ' . ' r’ '. . O D 0 . s'.‘ . ‘ n . .° . ' u .' '~ ' . ' e . . , . a. e I 1 . ' I I . ' . I O I . .' ' ' O O . . g g D . . o I I ‘ . . , . . ' -' ' 3“".1 I . . I If \ ' f . ‘r ' I ' \“ “.t ‘ . ' ' . . . :4 '0'“ . . o n I .O ' I I ‘ t . I s n .. I .’ I‘ '. '- . . I ’ ., z e . , 4 _ , ‘ : f y . _ l. s . ’e ‘1' '0' . ' I . ' '. , .. a .a‘ I. , s. _ o a , . ' _ . . I . a " .‘ 7- ‘ e . . - I v ," '. . . ' 4' r. s , I .0 .- v I I y . . , I' _’ “v 'a. - v s ’ ': ‘ ~ . ' .‘ .‘ J -' .- 1 .- ' .' ., 3 . - ‘- ' .’ o .' a '1 's E .' / ‘1 ." '. '. '. f e .’ q .9 '. ' ' l J J 1 c' . f ' J .- e .. ~' 4' " ‘. x ' " - ". o " y . ’ r " ¢ '.‘ . f u v .- 5 ‘v’ a V 5 ‘ - 5 I / e- V I -' I v. ~V. ‘1 . Q ‘4' v ~ K; v“ . s“ V. . K ' I ‘ \. ~‘ \‘ . s _ . “‘0‘ V»‘”. ' ‘A‘ ‘ wfvxméek‘t“ k Q. IIIII‘IWIIIIIIIIIIIIIIII 400 450 500 550 WAVELENGTH (nm) I. lamp spectra with 500 llllrll'Tfljl 600 650 700 nm cutoff filter 47 subtracted out. E. Resolution In multi-channel spectrophotometers the resolution is dependent on several things. The reciprocal linear disperison of the grating (32 nm/mm) is important as is the physical width of the array itself with respect to the spectral region covered. The spectral window is 400 nm. and the detector can be moved in the focal plane of the grating to obtain spectral information between 200 and 1000 nm. As mentioned previously. the entrance slit is interchangable. The resolution is ultimately limited by the width of the detector. However. with the smallest available slit width (25 um) the resolution is limited by the slit. This can be seen in Figures 5.8: 5.9 and 5.10. Figure 5.8 is a spectrum of a neon-filled hollow cathode lamp at a slit width of 100 um. Figures 5.9 and 5.10 were obtained with slit widths of 50 and 25 pm respectively. Resolution is calculated using the formula; L AA where baseline resolution is achieved. Calculated resolution values are 75.88 and 106 as the slit width is decreased. These were calculated using the identified lines on each spectrum. These values can be compared to the slit-limited bandpass values which are calculated using the following formula: INTENSITY (relative) 2500 2000 1500 1000 500 .441]llllilLJlllillLJilJlLllllllllLlllll‘lllllJLlll 48 594 ffizn N O U i 0 T T T T T T 1 T T I T T T T T T T T T 1 T T 1 T T T T T T 1 550 600 650 700 WAVELENGTH (nm) 5.8 Spectrum of a neon filled hollow cathode with a 100 pm slit width WHENSHY Oaknwe) 2000 r 1500 '7 w .l 1000 r .4 1 500 r 550 Fig 5.9 49 4J=réfl5 621 U UUU l T r T T l T T T l T T T T T I T T T I T T T T T T T T T 1 600 650 700 WAVELENGTH (nm) Spectrum of a neon filled hollow cathode with a 50 um slit width INTENSITY (relative) Fig. 50 l .J U . 1 \ M I _ J TTAT[WTTTTTVTTTTTTTTITTfFjTTTT‘I 550 600 650 700 WAVELENGTH (nm) 5.10 Spectrum of a neon filled hollow cathode with a 25 um slit width UT 5.. w "olyl where 8P (bandpass) is equal to the slit width multiplied by the reciprocal linear dispersion of the grating. The values calculated are 187. 386 and 795 for the 100. 50 and 25 um slits respectively. Further experiments were conducted to see if specific integration times degrade or improve resolution. Figures 5.11. 5.12 and 5.13 are spectra of the same hollow cathode lamp using a 25 Um slit width at 100. 500 and 1000 msec integration times respectively. Peak heights approach saturation as the integration time is lengthened. yet the resolution is not significantly decreased. The only perceptible difference is the loss of baseline resolution in the 1000 msec integration time spectrum. This is due to the broader base of the peaks. Also note the stray light (550 nm) in the longer integration time spectra. F. Saturation Effects The spectrum of a neon filled hollow cathode lamp in Figure 5.13 (integration time is 1000 msec and the slit width is 25 um) illustrates a very important characteristic of linear diode arrays. The array shows essentially no tendency to bloom. Blooming refers to the situation where a strong signal spreads to adJacent sensor elements. Therefore. even minor blooming can seriously degrade resolution and severely limit the use of the integrating Fig. msmr (relative) N O O 52 U ”UWUUUT 550 5.11 AAA T T T T T T T T T I I T T T T T T T T I T T T T I T T T T l 600 650 700 WAVELENGTH (nm) Spectrum of a neon filled hollow cathode at an integration time of 100 msec INTENSITY (relative) 53 I : n _: u ~ U " 5 ~ 2 “W u _ U V U U K Ij I T I T T T T I T T I T I I T T T I T T I T I T T I T] 550 500 650 700 WAVELENGTH (nm) 5.12 Spectrum 0? a neon Filled hollow cathode at an integration time o? 500 msec WHENSHY beknwe) 54 1000 900 800 700 600 500 400 300 200 W UUJ O TTTTITTTTITTTITITTTITTTITrTTIl 550 600 650 700 WAVELENGTH (nm) 100 lli;lLllLIlllLL_l_lllLiLllIlllllLlllLllllIlleJlJLLJ Fig. 5.13 Spectrum 0? a neon Filled hollow cathode at an integration time of 1.0 sec 55 capability of the array when measuring weak spectral lines in the presence of strong lines. Both silicon vidicons and charge coupled devices have problems with blooming. Thus with photodiode arrays the integration time can be used to increase sensitivity for weak lines. and intense lines will not interfere because of blooming even if they saturate the array. This would be of most benefit if the array were cooled. The only interference will be the broad base of strong lines which are present because of the finite slit width. Another important characteristic of electronic image sensors is lag. Lag refers to the image carry over from one scan to the next. In most electronic image sensors. the image cannot be completely read out in one scan. With the silicon vidicon for example. only 90% of the image may be erased on a readout cycle which leaves 10% to be carried over to the next scan. In general. lag is undesirable and is especially so if the detector is being used for time resolution studies. Linear diode arrays. however. do not. according to the literature (6). exhibit any lag. We have not tested this electrically since the present interface cannot store consecutive scans. G. Diode to Diode Sensitivity Variations This characteristic can be measured using the quartz halogen lamp as described in Chapter IV. A spectrum with 56 dark current subtracted is shown in Figure 5.15 (the integration time is 2.3 msec. the slit width is 100 11m). The variations seen in intensity along the curve (:12) are sensitivity variations from diode to diode along the array. The shape of the spectrum is not an accurate measure of the overall sensitivity variation across the array as this depends on a number of parameters such as the spectrum of the quartz halogen lamp. the polychromator throughput function and the spectral response of the photodiodes. With a calibrated source an overall spectral sensitivity function could be calculated and used to correct future scans. H. Dynamic Range The dynamic range of the array at a fixed integration time is an important characteristic. The dynamic range specified by Reticon is up to 4 orders of magnitude. The source used is the quartz halogen lamp at a 25 msec integration time and a 100 um slit width. A plot of percent transmittance at 650 nm versus peak amplitude is shown in Figure 5.16. The ZT axis was established using a series of neutral density filters of values 0.0132. 0.0525. 0.178. 0.302. 1.02. 2.95. 10.0. 19.5. 29.5. 41.7. 66.1 and 81.3 percent. The peak amplitude was recorded 5 times with each filter in place. The error bars show 11 standard deviation. A dynamic range of nearly 3 orders of magnitude is observed. INTENSITY (relative) 57 1.... .‘O“ ' . TITTITITIFITTfilllAT-TTIITTTIITITTTTTT—T—ITTIIIIITT] 250 300 350 400 450 500 550 600 650 700 Fig. WAVELENGTH (nm) 5.14 G.I. lamp spectrum showing sensitivity variations from diode to diode PEAK INTENSITY 58 O ITTIITTTTIIITTTWITTTTIITIITTIITTTTIIIIIIFTTTITTTW 0 1O 20 30 4O 50 60 7O 80 90 100 PERCENT TRANSMITTANCE Fig. 5.15 Peak intensity vs. Ztransmittance CHAPTER VI OPERATION AND APPLICATION OF THE INSTRUMENT A. Routine Operation of the Instrument The procedure for the operation of the LDA spectrophotometer is described in this section. The process begins by writing the desired subroutines using TECO (Text Editor and Corrector) on the 11/40 minicomputer. The operating system used on the 11/40 is RSXiiM V3.2. After creating the subroutine. it must be assembled. This is done by entering the following line: MAC (FILENAME).(FILENANE)=MACROS.GLOBALS.(FILENAME).END The first three letters. MAC. call the program MACRO which assembles the programs listed (after the equals sign) and creates the obJect and listing files (before the equals sign). A copy of the listing file serves as an invaluable debugging aid. The file HACROS is used to translate 8085 mnemonics into obJect code and also enter the library entries. GLOBALS contains pointers to all SLOPS routines which can be used by any or all user written routines. After GLOBALS all the user written routines are entered. END marks the end-of—library after the last user written subroutine. a When the MACRO assembly is complete. the obJect file is stripped of its files-11 format and converted to an unformatted obJect file. At this point the file is ready to 59 60 be transferred to a PERSCI formatted floppy disk. This is done using a program called PIPERSCI. The floppy disk is then compatible with the PERSCI dual disk drive on the LDA system. To run the subroutines written on the instrument. the following steps are taken: 1) Reset both the floppy disk and the microprocessor 2) Put the floppy disk in the drive 3) Depress the following keys on the SDK-BS: GO.A000.EXEC The microprocessor will then execute SLOPS which is written in the 2k of PROM which begins at memory address A000 (hexadecimal). The prompt SLOPS} will appear on the terminal screen. To load the user written routines. the disk handler must be called by typing DISK . The prompt Disk) will appear. Now the user written subroutines can be loaded into RAM by typing L (FILENAME) and responding Yes to the computer when it asks LOAD PROGRAM?. A control Z character exits from the Disk handler routine. The subroutine is now executed by entering its name followed by a carriage return. After execution is complete. the prompt SLOPS) will again appear. To obtain an absorbance curve of a sample. three spectra must be recorded; the reference. the dark current and the sample. with the current software this is done by Ientering the command CDATA. and storing the data when asked to do so. 61 The data files stored on the floppy disk can be read and transferred to a files—11 formatted file using the program PIPERSCI on the 11/40. The program ABSORB is now executed. This routine asks for the three input filenames and an output filename. The absorbance values at all 512 wavelength channels are sent to the output file in an ASCII data file.\ The routine MULPLT is the used to plot the data on the terminal or the lineprinter. B. Application of the Instrument One application of a linear diode array spectrophotometer is multicomponent analysis. It has been shown that multicomponent absorbance analysis is feasible for a few components in certain cases (28-29). Other applications include those in which the simultaneous ,recording of all wavelengths or the freedom from mechanical scanning systems are required. The first experiment conducted was the analysis of a series ' of potassium dichromate (KZCWZO7) solutions of varying concentration (5 ppm to 250 ppm). At a slit width of 100 pm and an integration time of 50 msec the observed absorbance at 350 nm with the most concentrated solution was 0.18. The expected absorbance is 1.87 at the 350 nm peak. The large error is due to the large amount of stray light present and the low source intensity at 350 nm. The experiment was repeated using a 500 nm cutoff filter (as 0 \ S0 shown in Fig. 5.7). The results improved somewhat to a maximum absorbance of 0.38 at 850 nm. This is still a very poor result for much the same reasons as before. The experiment was repeated a third time using a 400 nm cutoff filter and the resulting working curve is shown in Figure 6.1. The integration time was increased to 1000 msec because of the very low light levels in this region of the spectrum. The absorbance for the 250 ppm solution is now observed to be near 2.6. well above the expected 1.87. The working curve also has an increasing slope. These results are probably due to the non-linear dark current and light level response of the detector.' Upon cooling. it is expected that these results will improve. Also note that the standard deviations are very large which is due to the lower light level at high absorbance values. Another experiment conducted was the analysis of a series of potassium permanganate (KMnO4) solutions. This experiment was conducted to check the accuracy of the absorbance measurements at a longer wavelength where the source intensity is greater and the stray light is less of a problem. The resulting working curve is shown in Figure 6.2. The integration time is 85 msec and the slit width is 100 um. The results also show a positive deviation from Beer’s law which is due to the non-linear response of the detector. However. at this higher light level. the results are much closer to the expected value of 0.65 for the 42 ppm ABSORBANCE 31) 2f3 2() 1.5 1.0 (15 DJ) 63 300 T I I l I l 50 100 150 200 250 CONCENTRATION (ppm) Fig. 6.1 KZCrZO7 working curve ABSORBANCE 1.0 (19 C18 037 C16 C15 0J4 033 ()2 OJ OI) e4 O I I 5 I 10 Fig. I I I I l I I I 1 15 20 25 30 35 4O 45 50 55 60 IU CONCENTRAHON (ppno KMnOy working curve 65 solution and the standard deviations are much smaller. The utility of this spectrophotometer for precision absorbance measurements is yet to be determined as these preliminary results show that there are system improvements yet to be made. The first step to be taken is cooling the linear diode array. This is expected to nearly eliminate the non-linearities in the dark current and light level responses. In the ultra-violet region of the spectrum. it is necessary to increase the intensity for any reasonable results to be demonstrated. This can be easily done by a change in the source to a deuteriun lamp or similar type 0V source. Further studies using various bandpass and cutoff filters are expected to reduce the stray light so that absorbances near 2.5 can be measured accurately. APPENDIX A SELECTED PROGRAM LISTINGS i i 5 i 3 i 3 I SUBR SCAN: PUSH PUSH PUSH MVI STA LXI LXI SCAN1: INX MOV ANI MOV INX CALL JNC POP POP POP RET '0 O I 5 3 0 I i i i SUBR DISPLY: PUSH PUSH PUSH PUSH LXI MVI DISPLI: MOV DISPL2: MOV INX MOV ANI A. Microcomputer Software *************************§********************************* SCAN.MAC PETER AIELLO DEPT. OF CHEMISTRY MICHIGAN STATE UNIVERSITY EAST LANSING. MI 48824 *********************************************************** SCAN PSN iSAVE REGISTERS D 1 H 3 A.1 IBEGIN SCAN CSSCAN )AND COLLECT DATA D.DATA+2000 tINITIALIZE COUNTER H.DATA )GET BEGINNING ADDRESS H :00 TO TOP DATA BITS A.M iPUT IT ACCUMULATOR 17 :DELETE TOP FOUR BITS M.A :STORE TOP FOUR BITS ONLY H IINCREMENT TO iSKIP LON BITS DCMP :CHECK IF END OF DATA SCAN1 sCONTINUE H IRESTORE REGISTERS D 3 PSH 3 ******************************************%**************** DISPLY.MAC PETER AIELLO DEPT. OF CHEMISTRY MICHIGAN STATE UNIVERSITY EAST LANSING. MI 48824 ***************************************§******************* DISPLY PSN tSAVE REGISTERS B D H . H.DATA 300 TO START OF DATA 0.12 iSET BASE TO 12 B.C thITIALIZE COLUMN COUNTER E.M 3L0“ 8 DATA BITS TO E H iGET NEXT DATA A.M sTO ACCUMULATOR 017 iDELETE TOP 4 BITS 66 DISPLS: DISPL4: DISPL5: i i i i i i i i STORE: MOV INX CALL CALL LXI CALL JZ MVI CALL DCR JNZ CALL CALL JZ LDA CALL JZ LDA one LXI CALL CALL POP POP POP POP RET .ASCIZ SUBR PUSH PUSH PUSH PUSH CALL LXI 67 D.A H CVTEXT PRINTR D.DATA+2000 DCMP DISPL4 A.4O TTYOUT B DISPL2 CRLF CHARIN DISPLl KBDATA CHARIN DISPL3 KBDATA DISPLl H.DISPL5 PUSH2 PRINT H D B PS“ iFINIBH D.E PAIR IOF ALL 12 BITS ITO NEXT DATA POINT ICONVERT DATA TO BASE 10 iSEND POINT TO TERMINAL sCHECK IF END OF DATA .COMPARE TO H.L REGISTERS 51F END. PRINT END OF DATA IREADY TO PRINT TAB :PRINT IT iCOUNT COLUMN AS PRINTED 31F (12 COLUMNS. CONTINUE 31F 12. THEN CRLF sCHECK IF CHARACTER IIS TYPED )IF NO CHARACTER.TYPE sNEXT LINE IGET CHARACTER iAND THROW AHAY sCHECK FOR ANOTHER iCHARACTER )IF NO CHARACTER. IKEEP SEARCHING iGET CHARACTER iAND TYPE NEXT LINE :PRINT NEXT DATA IPRINT SEND IOF DATA TRESTORE REGISTERS IAND GO HOME IEND OF DATA/(CR) STORE.MAC PETER AIELLO *********************************************************** DEPT. OF CHEMISTRY MICHIGAN STATE UNIVERSITY EAST LANSING. STORE PSH B D H CLRDSK H.8TORE3 MI 48824 eaeeeeeeeeeeeeeeeeseeeeeeeeeeesseases**eeeeeeeeeeeeeeeeeeee TSAVE REGISTERS I i iCLEAR OUT ANYTHING IDISK IS DOING iASK FOR FILENAME STOREl: STORE2: STORES: STORE4: STORES: STORE6: CALL CALL MVI CALL CALL LDA LXI CALL MVI CALL CALL CPI JNZ CALL CPI JNZ LXI LXI MOV INX CALL MOV INX ANI CALL CALL JNC MVI CALL CALL CPI JNZ CALL CPI JNZ LXI CALL CALL POP POP POP POP RET .ASCIZ .ASCIZ LXI CALL CALL CALL 68 PUSH2 PRINT A.123 DDATUM TTYIN CHRNUM H.CHRBUF DDATA A.4 DCMD DINCHR 5 STORES DINCHR 4 STORES D.DATA+2000 H.DATA A.M H DDATUM A.M H 17 DDATUM DCMP STOREI A.4 DCMD DINCHR 6 STORES DINCHR 4 STORES H.STORE4 PUSH2 PRINT H D B PSN IPUSH ONTO USER STACK iAND SEND TO TERMINAL ILOAD SAVE COMMAND IAND SEND TO DISK sPUT FILENAME ON STACK iGET NUMBER OF iCHARACTERS TO SEND IAND BUFFER POINTER IAND SEND FILENAME TO DISK ILOAD AN EOT iAND SEND TO DISK IGET A CHARACTER IFROM THE DISK IIS IT AN ENG? iNO. START OVER IGET NEXT CHARACTER :IS IT AN EOT? iNO. START OVER .SET UP END OF .DATA LOCATOR .GO TO BEGINNING OF DATA .GET BOTTOM DATA BITS )INCREMENT TO HIGH 4 BITS ISEND TO DISK iAND GET THEM iGO TO NEXT DATA POINT iDELETE TOP 4 BITS iSEND TO DISK iCHECK IF END OF DATA iIF NOT LAST PT. CONTINUE ILOAD AN EOT TO iEND DATA SET IAND SEND IT iGET RETURN CHARACTER IIS IT AN ACK? sNO. START OVER iGET NEXT CHARACTER 518 IT AN EOT? 5ND. START OVER iSEND DATA STORE COMPLETE IPUSH ON STACK IAND SEND TO TERMINAL TRESTORE REGISTERS I iAND GO HOME /ENTER FILENAME: / /THERE HAS BEEN AN ERRORI/(CR) PETER AIELLO DEPT. OF CHEMISTRY MICHIGAN STATE UNIVERSITY EAST LANSING. *xeeseaeseeeeeeaeseassesseeaaaaaeeeessse**e******e***eeeeee CDATA.MAC . MI 48824 eeeeaeeaeesaeeeaeeeaas*eeaeaaeeeeaeaeeeeaeeeeeaa*eeeeeaeeea SUBR CDATA CDATA: PUSH H iSAVE REGISTERS PUSH PS” 3 LXI #H.CDATA2 iSEND INSTRUCTIONS CALL PUSH2 iPUSH ON STACK CALL PRINT TSEND TO TERMINAL CALL SCAN ICOLLECT DATA CALL DISPLY iDISPLY DATA ON TERMINAL LXI H.CDATA3 IASK TO STORE? CALL PUSH2 iPUSH ON STACK CALL PRINT iSEND TO TERMINAL CALL TTYIN IGET USER ANSWER LDA CHRBUF iAND LOAD ACCUMULATOR CPI 131 iYES? J2 CDATA4 iIF SO.CALL STORE CDATAl: POP PSN iRESTORE REGISTERS POP H 3 RET IAND GO HOME CDATA2: .ASCII {CRDITHIS PROGRAM IS NOW COLLECTING DATA/(CR) .ASCII /AND WILL DISPLAY DATA ON THE TERMINAL/(CR) .ASCIZ IANY KEY HALTS AND RESTARTS THE DISPLAY/(CR) CDATAS: .ASCIZ IDO YOU WANT TO STORE THIS DATA? [Y.NJ: I CDATA4: CALL STORE ISTORE THE DATA JMP CDATAI iAND END ROUTINE .aeeeaeaeaseeeeeeeeeaaeeeeaeeeeeaeeeeaeeeaeeeee*eeeeaaeeeeea SIGAVG.MAC 3 PETER AIELLO i DEPT. OF CHEMISTRY i MICHIGAN STATE UNIVERSITY EAST LANSING. MI 48824 *aeaaeaseseasseeaee*eeesaa*eeeseeeaeee*************eeeaesee SUBR SIGAVG SIGAVG: PUSH PSH iSAVE REGISTERS PUSH B 3 SIGAVI: SIGAV2: SIGAVS: SIGAV4: SIGAVS: PUSH PUSH LXI MVI LXI MOV INX CALL JNC LXI LXI MOV INX CALL JNC LXI CALL CALL MVI STA CALL CALL CALL JC XCHG SHLD SHLD CALL LXI LXI DAD MOV INX MOV PUSH LXI DAD MOV INX MOV XTHL DAD 70 D H DILDATA+2000 A10 H.LDATA M.A H DCMP SIGAVI H.HDATA D.HDATA+2000 M.A H DCMP SIGAV2 H.SIGAVS PUSH2 PRINT A.l2 RADIX TTYIN BRKDWN NUMBER SIGAV9 NSCAN COUNT SCAN B.O H.DATA B E.M H D.M D H.LDATA B E.M H D.M D iPREPARE TO ZERO MEMORY iPUT A 0 INTO ACCUMULATOR IGET ADDRESS WHERE iLOW 16 BITS iAND PUT O IN LOWER 8 BITS iNEXT ADDRESS i(D.E)-(H.L)CHECK 31F END OF DATA JAND CONTINUE iGET ADDRESS WHERE THIGH BITS STORED TPREPARE CHECK iSTORE O IN MEMORY iNEXT ADDRESS iCHECK IF END iOF DATA BLOCK iAND CONTINUE 3ASK FOR NUMBER OF SCANS iPUSH ON STACK )AND SEND TO TERMINAL iSET BASE TO 10 IAND STORE iGET RESPONSE FROM USER .PUT IN D.E REGISTER PAIR .IF ERROR. SEND MESSAGE .PUT IT H.L REGISTER PAIR .STORE IN MEMORY .FOR DIVISION .STORE IN MEMORY FOR COUNT IRUN A SCAN AND .STORE IN DATA .INDEX AND COUNTER .INITIALIZE .GO TO STARTING .ADDRESS OF DATA .INCREMENT To .ADDR. OF DATA TO AVG. .Low 8 BITS TO E .GET NEXT BITS .AND PUT IN D .AND SAVE LOW :16 BITS ON STACK .GET ADDRESS FOR .STORING LOW 16 BITS .ADDRESS OF AVERAGED DATA gLOW e BITS TO E .GET NEXT 8 BITS .AVERAGED DATA NOW IN D.E .H.L=DATA STACK=ADDRESS .AVERAGE IN H.L SIGAV6: SIGAV7: XCHG POP MOV DCX MOV JNC LXI DAD MOV INX MOV INX MOV DCX MOV INX INX MOV MOV LXI CALL JC LHLD DCX SHLD LXI CALL JC LXI PUSH LXI DAD MOV INX MOV PUSH LXI DAD MOV INX MOV LHLD XCHG POP 71 H M.D H M.E SIGAV6 H.HDATA B E.M H D.M D M.D H M.E B B D.E E.C H.2000 DCMP SIGAV5 COUNT H COUNT 0.0 DCMP SIGAV4 8.0 B H.LDATA B E.M H D.M D H.HDATA B E.M H D.M NSCAN IAVERAGE IN D.E TADDRESS IN H.L ISTDRE ANSWER 3(8 HIGH BITS) INEXT ADDRESS OF LOW BITS IAND STORE IIF CARRY.FILL sNEXT 16 BITS IGET HIGH 16 BITS ADDRESS IAND POINT TO CORRECT DATA IGET LOW 8 BITS TO E IINCREMENT ADDRESS THIGH 8 BITS TO D IINCREMENT REGISTER IPAIR D.E ISTORE ADDED DATA BITS I IAND NOW THE LON 8 BITS :INCREMENT COUNTER ITWICE(WORDS) IMOVE B AND C ITO D AND E ICHECK IF LAST IDATA PT. AVERAGED I(D.E)-(H.L) aGO TO AVERAGE NEXT POINT IGET NUMBER OF IDESIRED SCANS IDECREMENT BY 1 SCAN , ISTORE DECREMENTED COUNTER ICHECK IF LAST SCAN IAND SET ZERO FLAG IIF NOT LAST SCAN. ICONTINUE IINITIALZE COUNTER ITO STORE . IAND SAVE ON STACK IGET ADDRESS OF ILOW 16 AVG’D BITS IADD POINTER ILOW 8 BITS TO E iGET HIGH BITS )INTO THE D REGISTER ISTORE LOW 16 BITS .GET HIGH 16 BITS iADD POINTER iNEXT 8 BITS TO E .GET HIGH 8 BITS IIN TO THE D REGISTER .GET NUMBER OF SCANS IH.L=16 HIGHEST TBITS.D.E-DIVISOR TB.C=LOW 16 BITS SIGA13: SIGAVO: SIGAV9: SIGA10: SIGA11: SIGA12: CALL POP LXI DAD MOV INX MOV INX INX LXI CALL MOV MOV JC CALL LXI CALL CALL CALL LDA CPI JZ POP POP POP POP RET .ASCII .ASCII .ASCII .ASCIZ LXI CALL CALL JMP .ASCII .ASCIZ CALL JMP 72 DDIV D H.DATA D M.C H M.D D D H.2000 DCMP B.D C.E SIGAV7 DISPLY H.SIGAll PUSH2 PRINT TTYIN CHRBUF 131 SIGA12 H D B PSW IDIVIDE AND PUT IANSWER IN B.C IPOINTER INTO ID.E REGISTER PAIR IGET ADDRESS OF ISTORAGE SPACE IADD POINTER ISTORE LOW 8 BITS IINCREMENT ADDRESS ISTORE HIGH 8 BITS IINCREMENT POINTER ITWICE (WORDS) IINITIALIZE CHECK ICHECK IRESTORE POINTER ITO B.C REGISTER PAIR IIF NOT LAST POINT ICONTINUE IDISPLAY DATA ON TERMINAL )ASK TO STORE? ' IPUSH ON STACK ISEND TO TERMINAL IGET USER ANSWER IAND LOAD TO ACCUMULATOR IYES? IIF SO. CALL STORE IRESTORE REGISTERS I I IAND GO HOME ITHIS PROGRAM WILL TAKE THE NUMBER OF SCANS ENTERED./ STORE. ANY KEY HALTS AND RESTARTS (CR>/AVERAGE. / ,NOTE: THE DISPLAY/ AND DISPLAY THEM./ (CR)/PLEASE ENTER THE NUMBER OF SCANS TO BE AVERAGED: / H.SIGAIO PUSH2 PRINT SIGAV3 IPRINT AN ERROR MESSAGE IPUSH ON STACK IAND SEND TO TERMINAL IRETURN TO ASK AGAIN IPLEASE ENTER A DECIMAL NUMBER LESS THAT 64K/1 INS—::j[:>fi A.4 WD— "”7? FROM MICRO PROCESSOR All 4— D" SjDI ,+s 0— . D—fi>—5§ Alec. D COUNTERS FOR DMA TO INTERFACE RAM Fig. B. 2 COUNI ER RESET COUNTER CLOCK LATCH COUNTER OUTPUT ENABLE Counters and decoding for interface RAM 85 Ememmflt ufimemcum ~0Lu:0u mummeucH m.m .mau >mb30m5 JOthOU 0 39> AUISo xuoou Oluow DIES NONI O LIV OM92“. .50 I\_ coo: $538 . 183 Z 6.610 quw 183 COLIHV So 5538 AV Limo-u. “whzas . mu Ommefii ohm) 2w! _m d a: Sue I. .. .. _ 9: o _T enamel. ND 6 wmALmW G a . u v w I...» .00 .c- ..o no I Smog y AV rave. Ixuu. .uu XUOJU M. e 33% II. 865 0 Bow zw>wfiV HOSSBDOSd 083M W085 OMVOB BBHOdWV NODIIBCI W083 S6 ADC LATCHES AND CONVERTER TO INTERFACE RAM 0.. 234551 0o' 23 I 4 [ll warm ENABLE DUTCH OUTPUT ENABLE Lse use START ——-0 ADC CONVERSION 57“" com um PO 4" T———-—DVIOEO OUT UNI LA ADJUST MEL“ ZERO IO - l —+ ”rear 5 ANALOG 6N0 ‘1 o: g INTEGRATION CONTROL E! g 070543810 g (238 CS 74373 6N0 5° 98 go: In” —3 o: E” 74.53 GET 74l63 {3 Tc '" as HR m NITTc on M o m I- m ccpc - *5 1 +5 "W’ax<3 PE G Fig. 3.4 ADC and Integration control schematic diagrams 87 TIMING DIAGRAM I4l||2|3I4III l3|4III I3I4|I START —‘I VIDEO ' ' START CONV. I l E O C I LATCH ENABLEI l >x< 2:] E21 E23 DATA' fl LI— J )( COUNTER cuocx I] “—3 LF W U COUNTER M R :FIIFEIFD " Fig. 8.5 Control circu1trg timing diagram 88 ha Empmmflu .0. N0. 00. 00. wamemcum numog zqm x q omdom 24m v. w 30 V:N ¢:N .0. N0. ”0. V0. o§ 9+ .mfiu 9.4 n.< Na :4 0.4 APPENDIX C PERIPHERAL SCHEMATIC DIAGRAMS B9 l K PROM BOARD Fig. 0.2 1 k PROM board schematic diagram 90 empmmflu dongmgum upmon qum: m.o .mau memm 9.9+ Jazimww A n: Emma 4 we. a. 32.2%.? .0. o 9.. ice... on! _\ u m.— ¥.U v...U No mu am... Mr; <_0Nm at“ 9. Mn 3 «m uI»I_ no mm E I «a mu r. ox) I. 0+ APPENDIX D GRATING SPECIFICATIONS 91 GRATING SPECIFICATIONS -I +- Entrance Slit 3200 7040 SpOCIrum N IA . 96.9nm a --5.22|7° Ia - I05.44nm Rama-HQ??? IB - I05.49nm BwooHIZJBZ" I3 . I06.92nm %m-*l5.724° I9 - IOZ48nm Amour/.590“ Fig. 0.1 Grating specifications REFERENCES florists» 10. 11. 12. 13. 14. 15. 16. 17. 18. REFERENCES R. M. Barnes. Anal. Chem.. 41. 122R (1972) Hillard. Merritt and Dean. "Instrumental Methods of Analysis". Fifth Ed.. D. Van Nostrand Company. New York. 413 (1974). R. E. Santini. M. H. Milano and H. L. Pardua. Anal. Chem.. 53. 915A (1973). 0. Horlick and E. G. Codding. Anal. Chem.. 42. 1490. (1973). “— Y. Talmi. Am. Lab.. March. 79 (1978). Horlick. Appl. Spectrosc.. g9. 113 (1976). Talmi. Anal. Chem.. 42. 658A (1975). Talmi. Anal. Chem.. 22' 697A (1975). 3 7‘ :< 9 . N. Busch and G. H. Morrison. Anal. Chem.. 43. 712A (1973). “’ J. D. Ninarordnar. J. J. Fitzgerald and N. Dmanatto. Appl. Spectrosc.. 23. 369 (1975). R. P. Coomeq. G. D. Boutiliar and J. D. Hinafordner. Anal. Chem.. 33. 1048 (1977). D. F. Brost. B. Mallog and K. w. Butch. Anal. Cham.. 42. 2280 (1977). H. L. Folkel and H. L. Pardue. Anal. Chem.. §9. 602 (i978). ‘— N. 0. Howell and G. H. Morrison. Anal. Chem.. 49. 106 (1977). ‘- H. L. Felkol and H. L. Pardue. Anal. Chem.. 423 1112 (1977). F. S. Chaung. D. F. S. Natusch and K. R. D'Keefe. Anal. Chem.. 29. 525 (1978). 0. Horlick and E. G. Codding. Appl. Spectrosc.. 29. 167 (i975). —— K. M. Aldous. D. 0. Mitchell and K. w. Jackson. Anal. 92 19. 20. 21. 22. 23. 24. 25. 26. 27. 2B. 29. 30. 31. 32. 33. 34. 35. 93 T. L. Chester. H. Haraguchi. D. D. Knapp. J. D. Hessman and J. D. Ninefordner. Appl. Spectrosc.. 29. 410 (1976). "— T. E. Cook. M. J. Milano and H. L. Pardue. Clin. Chem. 29. 1422 (1974). E. G. Codding and G. Horlick. Spectrosc. Lett.. Z. 33 (1974). T T. E. Edmonds and G. Horlick. Appl. Spgctrosc.. 31. 536 (1977). “— M. Franklin. C. Baber and S. R. Koirtgohann. Spectrochim. Acta.. Part B. 21. 589 (1976). K. R. Bettg and G. Horlick. Appl. Spectrosc.. 32. 31 (197B). ‘— G. Horlick. E. G. Codding and S. T. Leung. Appl. Spectrosc.. 23. 48 (1975). K. N. Busch. N. G. Howell and G. H. Morrison. Anal. Chem.. 36. 2074 (1974). N. G. Howell. J. D. GanJei and G. H. Morrison. Anal. £233;' 42. 319 (1976). A. E. McDowell. R. S. Hanner and H. L. Pardue. Clin. Chem.. 22. 1862 (1976). M. J. Milano and K. Y. Kim. Anal. Chem.. 52. 555 (1977). -— A. E. McDowell and H. L. Pardue. Anal. Chem..'fl3. 1171 (1977). “7 R. E. Desseu. N. D. Reynolds. w. G. Nunn. C. A. Titus. and G. F. Moler. Clin. Chem.. 22. 1472 (1976). J. R. Jadamec. N. A. Sauer and Y” Talmi. Anal. Chem.. 43. 1316 (1972). R. P. Cooneq. T. Vo-Dinh and J. D. wineFordner. Anal. Chim. Acta.. 22. 9 (1977). 5.0.0. RETICDN. S—Series Solid State LDA. Brochure (1978) Bruce Newcome. Dept. of Chemistry. Michigan State University (1980). 36. 37. 38. 39. 40. 94 Hugh Gregg. Dept. of Chemistry. Michigan State University (1980). Product Specifications on Model 1070 Intelligent Diskette Controller. Persci. Marina Del Rey. CA. (1976). T. A. Neiman. Ph.D. Thesis. Michigan State University. (1975). J. E. Hornshuh. Ph.D. Thesis. Michigan State University. (1978). T. B. McCord and M. J. Frankston. Appl. Optics. 13. 1437 (1978). ‘— E. M. Carlson. Ph D. Thesis. Michigan State University. (1978). ARIES IIIIIIIIIIII