Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The development of a completely automated oxygen isotope mass spectrometer Ahern, Timothy Keith 1980

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1980_A1 A48.pdf [ 6.85MB ]
Metadata
JSON: 831-1.0052992.json
JSON-LD: 831-1.0052992-ld.json
RDF/XML (Pretty): 831-1.0052992-rdf.xml
RDF/JSON: 831-1.0052992-rdf.json
Turtle: 831-1.0052992-turtle.txt
N-Triples: 831-1.0052992-rdf-ntriples.txt
Original Record: 831-1.0052992-source.json
Full Text
831-1.0052992-fulltext.txt
Citation
831-1.0052992.ris

Full Text

THE DEVELOPMENT OF A COMPLETELY AUTOMATED OXYGEN ISOTOPE MASS SPECTBOMETEB by TIMOTHY KEITH AHEBB B.S., Whitsorth College, 1972 H.Sc., University of B r i t i s h Columbia, 1975 A THESIS SUBMITTED IN PARTIAL 'FULFILLMENT OF THE SEQUIBEMENTS FOR THE DEGBEE OF DOCTOB OF PHILOSOPHY in THE FACULTY OF GBADUATE STUDIES (Department of Geophysics and Astronomy) We accept t h i s thesis as conforming to the required standard The University Of B r i t i s h Columbia March, 19 8 0 (c) Timothy Keith Ahern In presenting th i s thesis in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Univers ity of B r i t i s h Columbia, I agree that the Library shal l make i t f ree ly avai lable for reference and study. I further agree that permission for extensive copying of th i s thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It i s understood,that copying or publ icat ion of th i s thesis for f inanc ia l gain shal l not be allowed without my written permission. Department Of Geophysics and Astronomy The Univers ity of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 D a t e A p r i l 27, 1980 Abstract k completely automated mass spectrometer system has been developed to measure the oxygen isotope:ratio of carbon dioxide samples. The system has been shown to have a precision of 0.03°/oo, which i s comparable to that quoted for any other system i n the world._ In addition, the f a c i l i t y i s capable of analyzing over one hundred samples per day. The system uses an Interdata minicomputer as the primary c o n t r o l l e r . The:minicomputer monitors the quality of analyses, on-line, and thereby insures that a l l DEL values are measured to at lea s t 0.04°/oo. Host of the sophistication resides i n i n t e l l i g e n t c o n t r o l l e r s within the mass spectrometer console.. This design gives a technician considerable power when operating the system i n a manual mode. The i n t e l l i g e n c e of theisystem i s contained within hardware c i r c u i t s , software within the minicomputer and firmware written f o r a Motorola 6802 microprocessor.. i major contribution of t h i s thesis has been the.design and i n s t a l l a t i o n of an automated mass spectrometer i n l e t system. . 4 microprocessor based i n l e t system c o n t r o l l e r maximizes the throughput of carbon dioxide samples within the i n l e t system.. The i n l s t system normally contains four d i f f e r e n t aliquots of carbon dioxide and introduces these samples to the mass spectrometer, in proper sequence, through a single mass spectrometer admittance leak.. The system has been used in the analysis of 1 11 samples of ice taken from the Steele Glacier, Yukon T e r r i t o r y . , The samples taken from a v e r t i c a l borehole, displayed a sawtooth variation of the oxygen isotope r a t i o with depth. The data have been explained by a physical model described in an appendix to t h i s thesis.. I f our inter p r e t a t i o n i s correct, the i s o t o p i c variations have recorded at le a s t four surges of the.Steele Glacier. . i v TABLE OF CONTENTS Abstract ................................... ,«,.... ..... , i i L i s t 3f Tables . . ...... ........<........................ v i i L i s t Of Figures .................. .... . . . ... . ... ..... y i i i Acknowledgments . ..... . . .. ............. . • .............. . x I. Introduction ................................... . i . . . 1 1.1 The Measurement Of The Oxygen Isotope Ratio In Hater ................ ............... • « .. 1 1.2 The History Of Oxygen Isotope Measurements At The University Of B r i t i s h Columbia . . . . . . o . . . . . . : 3 1.3 Objectives Of This Thesis Project 6 II. , Improvements i n T r a d i t i o n a l Mass Spectrometry ....,12 2.1 A New Technique of Gas Source Construction .... 12 2.2 The.Dual Beam Ion Collector ................... 20 2; 3 Direct Regulation Of The Magnetic F i e l d <i....... 35 III. , Innovations In D i g i t a l l y Controlled Mass Spactrometry ......................................... 43 3.1 A Brief Description Of Peripheral i n t e r f a c i n g To Interdata Minicomputers ...................... 43 3.2 The D i g i t a l Magnet Controller ........... ••••••• _ 45 3.3 D i g i t a l Compensation Of Preamplifier Baselines 52 3.4 The Design Of A Multiplexed Analog To D i g i t a l Converter _ 59 3.5 The General D i g i t a l Input/Output Interface Module ...,..,.,..,..,........,....,.,,,,«...., ,...,64 3.6 High Voltage Beam Suppression C i r c u i t r y ....... 69 IV. . The Dual Sided I n l e t System 71 V 4.1 General Description ...................... ...... ,7 1 4.2 The Toepler Pump Controller ................... 79 4.2.1 The Toepler Pump Analog Computer w.i.^...i...81 4.2.2 The Toepler Pump D i g i t a l Logic Board 86 4.2.3a The Toepler Pump Driver Board .......... 1. 91 4.2.3b The Mercury Level Centering C i r c u i t r y ...97 4.2.4 The Design Of A Multiplexed D i g i t a l Voltmeter .98 4.3 The Microprocessor Based I n l e t System Controller .....................,..............,. 102 4.3.1 General Description 102 4.3.2 Standby, Manual And Universal Pump Control Modes ....... -................ 109 4.3.3 Standard Load Control Mode ....,.......... 110 4.3.4 The GO Control Mode .................... 112 4.3.5 The Inlet System Pump Routines ...........119 V. The Automated System .............................. 121 5.1 Preliminary Bemarks ........................... 12 1 5.2 Advantages Of An Automated Oxygen Isotope Mass Spectrometer ....................................122 5.3 System Description ............................124 5.4 The Bole Of The Interdata Computer ............131 5.4.1 On-line Data Reduction .,..131 5.4.2 The Minicomputer As The Central Controller .135 5.4.3 The Peak Centering Algorithm 138 5.5 The V e r s a t i l i t y Of The System i. .... i 14 1 v i 5.7 The Performance Of The Automated System ••••••• 144 L i s t Of Works Consulted .150 Appendix I Interdata Assembly Language Control Program 155 Appendix II Motorola Assembly Language Inlet System Control Program ........ .,. . . . . . . . . . . . . . . . . . . . . . . . . . . J 56 APPENDIX I I I Isotopic Evidence Of Surges Of The Steele Glacier, Yukon Territory .......................... .157 v i i LIST OF TABLES 2.1 Comparison Of Focal Points Determined By Herzog And Cartan Methods 23 2.2 Focal Point Locations For A Specific Asymmetric Configuration Of Source And Magnet 30 4.1 The Possible Combinations Of Peak Height Com parisons .85 5.1 Isotopic Values Obtained From A Suite Of Linearly Varying Water Samples ..............................144 5i2 The Results Of The Automated Analysis Of 22 Iso t o p i c a l l y Identical C02 Samples .................145 5.3 The Results Obtained For The Reanalysis Of Selected Samples From The Steele Glacier ........... 148 v i i i LIST OF FIGURES 2.1 Side View Of MACOH Gas Source .................... 14 2 . 2 The Complete Gas Source . . . . . . . . . . . 1 8 2 . 3 A Comparison Of Two Methods Of Focal Point Determination . ........................,• ., 25 2 . 4 The E f f e c t Obtained By Moving The Magnet In A Horizontal Direction . . . . . . . . . . . . . . . . . . . . . . . 2 8 2 . 5 The Dual Beam Collector Assembly . . . . 3 1 2 . 6 C i r c u i t r y Used For Direct Regulation Of Magnetic F i e l d . . 37 2 . 7 The S t a b i l i t y Of Measured DEL Values Using Direct F i e l d Regulation .............. . . . . . , . . . . . . • , . , . . . . . , . 41 3.1 D i g i t a l Magnet Controller C i r c u i t r y .............. 47 3 . 2 D i g i t a l Baseline Compensation C i r c u i t r y . . . . . . . . . . 5 4 3 . 3 A. »6 Channel Multiplexed Analog To D i g i t a l Con verter ....................... . . . . . . . . . . . . . . . . . . . 6 1 3 . 4 The General input/Output Logic Module . . . . . . . . . . . . 66 4.1 The Dual Sided I n l e t System . . . . 7 2 4 . 2 Cross-sectional View Of The Toepler Pumps Designed For The Automated System 76 4 . 3 The Analog Computer C i r c u i t Of The Toepler Pump Con t r o l l er ......................... . , 82 4 . 4 The TPC D i g i t a l Logic Control Board .............. 87 4 . 5 Control Valve Configuration For The Toepler Pumps 92 4 . 6 The Toepler Pump Driver Board ....................94 4 . 7 The Toepler Pump Multiplexed D i g i t a l Voltmeter C i r c u i t 100 4.8 The B l o c k , Diagram Of The M i c r o p r o c e s s o r Based I n l e t System C o n t r o l l e r . . . . . . . . . . . . . . . . . . . . . . . . o . . . 1 0 5 4 . 9 a A. F l o w c h a r t Of The F u n c t i o n i n g Of The GO C o n t r o l Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 4 4.9b The F l o w c h a r t Showing The I n t e r r u p t S e t v i c e R o u t i n e D e v i s e d F o r The I n l e t System C o n t r o l l e r . i t . . 1 1 7 5.1 B l o c k Diagram Of The Automated Oxygen I s o t o p e Mass S p a c t r o m e t e r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 5 5.2 The Method Of L i n e a r I n t e r p o l a t i o n Used In The Data R e d u c t i o n 132 A3. 1 Map Of The S t e e l e G l a c i e r Showing Sample L o c a t i o n s 159 A3.2 The Measured Oxygen I s o t o p e V a r i a t i o n As A F u n c t i o n Of Depth Taken From Thermal D r i l l Hole Number One . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 A3.3 I s o t o p i c P a t t e r n s E x p e c t e d From A G l a c i e r In E q u i l i b r i u m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 A3.4 The I n i t i a l P o s i t i o n Of Two I s o d e l L i n e s In The S t e e l e G l a c i e r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 A3.5 The Shape Of The I s o d e i L i n e s A f t e r A S i n g l e Surge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 1 A3.6 I s o d e l P a t t e r n Formed A f t e r Two S u r g e s For An I n i t i a l l y N o n - v e r t i c a l I s o d e l L i n e . . . . . . . . . . . . . . . . . 174 A3.7 I s o d e l P a t t e r n Formed A f t e r Ten Surges F o r An I n i t i a l l y N o n - v e r t i c a l I s o d e l L i n e . . . . . . . . ; . . . . . . . . 176 A3.8 P o s i t i o n Of Two I s o d e l L i n e s O b t a i n e d A f t e r Ten Surges Of A G l a c i e r E x h i b i t i n g E x t e n d i n g Flow . . . . . . 1 7 9 X acknowledgments I wish to thank R.D. Meldrum and K.p., Schreiber for th e i r cooperation i n thiis thesis project. . A. project involving the design and construction of so many di f f e r e n t components can only be successful i f one has the assistance of highly competent in d i v i d u a l s such as they. ., R.D. Meldrum assisted in the design and i n s t a l l a t i o n of several of the elect r o n i c supplies as well as providing much needed advice and encouragement., K.D., Schreiber has spent endless hours constructing many of the i n d i v i d u a l mass spectrometer components.„ His competence and friendship w i l l not be forgotten. H.W. Verwoerd constructed much of the: c i r c u i t r y required for t h i s thesis project. His expertise and advice helped me through many of the d i f f i c u l t i e s encountered. Capable tec h n i c a l assistance. was also provided by P. Michalow and W. Siep on several occasions. . I would l i k e to thank Dr. G.K.C. Clarke:for t h e f i e l d support he provided while c o l l e c t i n g the Steele:Glacier samples., His tuna bisque w i l l be remembered.. I would also l i k e to express my appreciation to S. C o l l i n s and P.. Cary for a s s i s t i n g me i n the: f i e l d c o l l e c t i o n of the Steele Glacier samples.. Sany f r u i t f u l discussions were held with E. Waddington and Dr. Clarke concerning the interpretation of the Steele Glacier r e s u l t s . I would l i k e to thank Jacquie Gates for entering t h i s x i thesis i n t o the computer.. Her e f f o r t s made the.production of t h i s thesis much easier.. The Computing Centre of the University of B r i t i s h Columbia supported the production of t h i s t h e s i s . I would l i k e to acknowledge f i n a n c i a l support received in the form of three National Research Council scholarships. I received additional funding from research assistantships from grants made to my supervisor. Dr. R.D» - Russell.. The entire project was funded by National Research Council grant number A0720 given to Dr. .. Russell. . The purchase of several mass spectrometer components was made possible through National Research Council grant number E3551.a The Steele Glacier f i e l d work was financed by a University of B r i t i s h Columbia A r c t i c and Alpine grant to Dr. Russell. 1 thank him for the considerable f i n a n c i a l support he has given t h i s project.. F i n a l l y I would l i k e to acknowledge the unlimited help and encouragement given to me by my wife, Rowena. Not only did she perform the Steele Glacier analyses but she also devoted a tremendous amount of energy to drafting and l e t t e r i n g a l l the figures contained in t h i s thesis. . No words can express my gratitude to her. It i s with great happiness that I dedicate t h i s thesis to her. 1 I. INTRODUCTION 1.1 The Measurement of the Oxygen Isotope Ratio i n Water The measurement of the.variation i n the.isotopic r a t i o of oxygen i n water i s routinely carried out by approximately 50 laboratories around the world.. Most of these f a c i l i t i e s have adopted the. procedure described by Epstein and Mayeda (1953) which involves i s o t o p i c a l l y e q u i l i b r a t i n g the water samples with carbon dioxide and subsequently analyzing the carbon dioxide i n a mass spectrometer. Mass spectrometers used to measure oxygen isotope ratios in carbon dioxide are usually based on the design of A, ,0. Nier (1947). Modifications to t h i s basic design are usually those described by McKinney et a l . (1950) and Nier et a l . .(1947),. These modifications enable very small i s o t o p i c differences to be detected in a sample of carbon dioxide by using a n u l l method of ion current comparison and a dual sided i n l e t system which allows rapid intercomparison between a working standard and the unknown carbon dioxide. Stable isotope r a t i o s are generally reported i n terms of a DEL function: DEL (x/s) = (Rx/Rs- 1) • 10 3 [1-1] where Rx and Rs refer to the r a t i o s of the i n t e n s i t i e s of 2 the mass 46 and mass 44 ion beams for the carbon dioxide equilibrated with the unknown water and the standard carbon dioxide respectively. The precision of analyses i s generally deemed acceptable i f analyses can be reproduced to within 0.1°/oo. To f a c i l i t a t e intercomparison of results between laboratories the standard used i n equation 1.1 i s i d e a l l y Standard Mean Ocean Water (SMOW), Craig (1961) . No samples of t h i s true SMOW are available and the standard generally used today i s Vienna-SMOW (V-SMOW) whose:0* 8/0 1 6 absolute r a t i o has been determined by Baertschi (1976) to be (2005.20 + 0.45) x 10^6. I t follows by d e f i n i t i o n that DEL (V-SMOW) = 0.00. Three other world standards have been d i s t r i b u t e d in the. past by the International Atomic Energy Agency (I.A.E.A.).. These are NBS1, NBS1A and SLAP (Standard Light Antarctic P r e c i p i t a t i o n ) . At the Consultants' Meeting on Stable. Isotope Standards and In t e r c a l i b r a t i o n i n Hydrology and Geochemistry held i n Vienna, Austria i n September, 1976, i t was agreed that laboratories should l i n e a r l y scale t h e i r values such that each laboratory's corrected value of DEL(SLAP/V-SMOW) i s -55.5°/oo (Gonf i a n t i n i , 1977).. I t should be noted that DEL(SMOW/V-SMOW) = +0.05 following the suggested procedure and i n most cases the . difference w i l l remain undetected. v Since the above discrepancy introduces a systematic error, and one i s normally interested i n r e l a t i v e i s o t o p i c values, the above convention i s e n t i r e l y acceptable. , 3 1.2 The History of Oxygen Isotope Measurements at the University of B r i t i s h Columbia The Department of Geophysics and Astronomy i n i t i a t e d the conversion of one of i t s three mass spectrometers to an oxygen isotope f a c i l i t y in the summer of 1972. Procedures had been refined by November, 1973 to a point where analyses were reproducible to at least 0.2 parts per thousand.. By November, 1974 an i n i t i a l project to measure the oxygen isotope composition of 405 g l a c i e r and snowfield samples from Devon Island was completed for the Polar Continental Shelf Project (Russell et a l . 1974). In January of 1974 t h i s writer, with the:assistance of Dr. . K..E. . West, i n i t i a t e d a study of water percolation through sub-zero snowpacks as a f i e l d project associated with a masters thesis (Ahern, 1975a). By the conclusion of t h i s project the laboratory had acquired a s i g n i f i c a n t amount of experience i n the f i e l d of oxygen isotopes, i n p a r t i c u l a r the mass spectrometric analysis of the oxygen isotope r a t i o i n water, and had also gained l i m i t e d experience i n the analysis of carbonate samples, (Fink, 1974 ; Ahern et a l . 1980) . By the conclusion of the above studies, the laboratory was routinely analyzing samples with a precision of 0.14°/oo. However, the mass spectrometer was plagued with many d i f f i c u l t i e s . I t had marginal s t a b i l i t y , i t had a very inadequate vacuum system and i t was extremely slow and awkward in sample handling., On separate occasions i n 1975, 4 Dr. _ S. . D. . Russell and T. . K. . Ahern v i s i t e d the Geophysical Isotope Laboratory i n Copenhagen, Denmark and saw the f a c i l i t y that Dr..W..Dansgaard had established.. We were both very impressed with the system and f e l t that the establishment of a s i m i l a r f a c i l i t y in Canada was both appropriate and needed. The Copenhagen f a c i l i t y i s capable of analyzing up to 256 samples d a i l y with a precision of about 0.05 DEL (Gundestrup, personal communication).. Laboratories that measure oxygen isotope r a t i o s t y p i c a l l y measure 16 to 20 samples a day. In 1975 the mass spectrometry laboratory at the University of B r i t i s h Columbia could only produce about 8 analyses per 8 hour day, when no problems occurred. The system was also plagued with electronic and mechanical problems and c l e a r l y had to be improved i f analyses were to contin ue. One important role of the.mass spectrometry laboratory at the University of B r i t i s h Columbia i s i t s function as a mass spectrometer development f a c i l i t y where new techniques and state of the art instrumentation have t r a d i t i o n a l l y been introduced.. This function has altered recently with the Department of Geological Sciences playing a larger role i n running the s o l i d source mass spectrometer used in lead analyses. Their p r i n c i p a l i n t e r e s t i s i n the application of the measurements to geological problems. The oxygen isotope f a c i l i t y has, u n t i l recently, continued to occupy a developmental position i n the 5 laboratory although several hundred analyses were performed by R. Ahern i n an 8 month period in 1978., More recently J . . Eng completed measurements on additional samples. . In the period 1976 to 1978 four oxygen isotope studies were undertaken in the following areas; isotopic fractionation i n pingo growth, (Ahern, 1975b), compositional changes i n p r e c i p i t a t i o n samples collected across B r i t i s h Columbia, the use of oxygen isotopes as an environmer-?.l tracer i n animals (Ahern et a l . 1980) and a major study of oxygen isotope variations i n a surging type g l a c i e r to be. discussed l a t e r in t h i s thesis. W. F. Slawson i n i t i a t e d a study of the i s o t o p i c composition of groundwater on the University of B r i t i s h Columbia campus in May, 1977. Analyses of approximately 150 samples have been completed and are awaiting int e r p r e t a t i o n . The above summarizes most of the major a c t i v i t y in the use of oxygen isotopes over the past several years. However, the bulk of my e f f o r t s have been directed toward a complete redesigning of the mass spectrometer f a c i l i t y by improving the components used i n the mass spectrometer as well as designing d i g i t a l c o n t r o l l e r s and implementing techniques to increase sample throughput and heighten the r e l i a b i l i t y of analyses,. The remainder of t h i s thesis i s devoted to a discussion of the.system which has been developed in order to obtain oxygen analyses of both high quality and number.. 6 1.3 O b j e c t i v e s of t h i s T h e s i s P r o j e c t The l a r g e expanses of Canada t h a t l i e . w i t h i n the A r c t i c make s t u d i e s of snow and i c e most r e l e v a n t * , I t i s t h e r e f o r e r e a s o n a b l e to have a C a n a d i a n f a c i l i t y t h a t i s c a p a b l e o f p e r f o r m i n g l a r g e numbers of h i g h p r e c i s i o n a n a l y s e s of t h e oxygen i s o t o p e r a t i o i n w a t e r . . The major o b j e c t i v e o f t h i s t h e s i s p r o j e c t was t o d e s i g n and c o n s t r u c t an automated mass s p e c t r o m e t e r t h a t had the c a p a b i l i t y of p r e c i s e l y m e a s u r i n g the r a t i o o f 0 1 8 t o 0 1 6 i n c a r b o n d i o x i d e s a m p l e s . . A t t a i n m e n t o f t h i s g o a l has been d e m o n s t r a t e d by r e p r o d u c i n g r e s u l t s t h a t had p r e v i o u s l y been o b t a i n e d i n a manner r e q u i r i n g c o n s i d e r a b l e o p e r a t o r i n t e r v e n t i o n . The samples i n g u e s t i o n a r e those o b t a i n e d d u r i n g a f i e l d s t u d y of the S t e e l e G l a c i e r * Yukon T e r r i t o r y . . The s t a t e d o b j e c t i v e was a c h i e v e d by making advances i n both a r e a s t h a t a r e t r a d i t i o n a l mass s p e c t r o m e t r y and a l s o by a p p l i c a t i o n o f d i g i t a l e l e c t r o n i c s i n a r e a s of mass s p e c t r o m e t r y t h a t had p r e v i o u s l y been i g n o r e d o r a c c o m p l i s h e d manually* The i n i t i a l t a s k i n t h i s t h e s i s p r o j e c t i n v o l v e d a c l e a r f o r m u l a t i o n o f the method t h a t c o u l d p r o v i d e the needed i mprove me nts. D u r i n g my M . S c . t h e s i s p r o j e c t , a g r e a t d e a l o f i n f o r m a t i o n was o b t a i n e d which a l l o w e d t h e d e t e r m i n a t i o n o f t h e p i t f a l l s t h a t e x i s t i n manual a n a l y s e s . I t i s my g e n e r a l b e l i e f t h a t the key t o s u c c e s s i n s t a b l e i s o t o p e measurements i s c o n s i s t e n c y i n the t e c h n i q u e used t o 7 o b t a i n the r e s u l t s . The method o f a n a l y s i s w i l l o f t e n g r e a t l y d i m i n i s h e r r o r s i n t e c h n i q u e as l o n g as b o t h the s t a n d a r d s and unknowns have r e c e i v e d i d e n t i c a l t r e a t m e n t . T h i s f a c t l e a d s one t o r e c o g n i z e the g r e a t advantage ; t h a t an automated system has o v e r manual o p e r a t i o n s s i n c e t h e f o r m e r t e n d s t o be v e r y c o n s i s t e n t . A n o t h e r a d v a n t a g e : t h a t automated mass s p e c t r o m e t r y o f t e n p o s s e s s e s i s the speed with which most f u n c t i o n s can be performed;. _ E x p e r i e n c e w i t h the i n i t i a l oxygen i s o t o p e f a c i l i t y i n d i c a t e d t h a t a r e a s o f t r a d i t i o n a l mass s p e c t r o m e t r y which needed improvement were: 1. The d u a l beam c o l l e c t o r assembly s h o u l d be r e d e s i g n e d so t h a t b o t h beams c o u l d be f o c u s s e d on t h e i r r e s p e c t i v e d e f i n i n g s l i t s . . The r e s u l t i n g improvements i n peak shape would i n c r e a s e s t a b i l i t y . 2. , The gas s o u r c e had t o be m o d i f i e d t o i n c r e a s e s e n s i t i v i t y which would a l l o w lower f i l a m e n t t e m p e r a t u r e s and r e s u l t i n l o n g e r component l i f e and g r e a t e r s t a b i l i t y * . 3. . The magnetic c u r r e n t s u p p l y had t o be m o d i f i e d so t h a t i t was r e g u l a t e d d i r e c t l y by t h e f i e l d r a t h e r t h a n by t h e c u r r e n t t h r o u g h the magnet c o i l s . Long and s h o r t term magnetic f i e l d v a r i a t i o n s c o u l d be e l i m i n a t e d . . 8 In addition to these needed improvements, i t was f e l t that a modification of the i n l e t system that would allow both standard and unknown carbon dioxide to enter the mass spectrometer through a single leak could eliminate many d i f f i c u l t i e s others have had i n matching leak c h a r a c t e r i s t i c s (Coleman and Gray, 1972) or in applying leak corrections as indicated by Begbie et al._(1972). A study by Russell and Ahern (1977) provided the framework fo r an innovative technique of sample l i n e side selection., This allowed a single molecular leak to be used instead of the normal design u t i l i z i n g two magnetically controlled high vacuum valves. The application of d i g i t a l control techniques to the gas source mass spectrometer i s an extremely s i g n i f i c a n t contribution of t h i s thesis. Considerable appreciation has been gained for the contributions a s k i l l e d operator can make toward a successful f a c i l i t y . To replace his a b i l i t y with automated c o n t r o l l e r s i s not a t r i v i a l task., Recent advances i n d i g i t a l e lectronics culminating i n microprocessors allow extremely powerful control techniques to be applied to mass spectrometry once the problem areas have been i d e n t i f i e d . . There were no less than f i v e problem areas that have been solved through the application of d i g i t a l control techniques. These areas were either i d e n t i f i e d i n the previously mentioned thesis (Ahern, 1975a), or were recognized in i n i t i a l tests of the new system. The 9 following problems have been addressed by d i g i t a l techniques; Long term d r i f t of either the magnetic f i e l d or high voltage can be compensated by d i g i t a l l y c o n t r o l l i n g the magnetic f i e l d . . Peak centering, beam supression and peak hopping algorithms have been implemented on an Interdata d i g i t a l computer.. 2. „. Baseline d r i f t of the two parametric amplifiers described by Russell and ahern (1974) can cause s i g n i f i c a n t errors i n the measured DEL values. Two d i g i t a l units, which function as analog memories, have been inserted i n a feedback loop which eliminates errors caused by baseline o f f s e t s . . 3., Coleman and Gray (1972) have shown that differences i n operating pressure between the analysis of the standard carbon dioxide, and the unknown can dramatically affect the measured DEL value. Mercury f i l l e d Toepler pumps have been included i n the new gas i n l e t system and they are i n t e l l i g e n t l y controlled by using integrated c i r c u i t s from the Low Power Schottky TTL l o g i c family. , 4. , Manually operated stopcocks i n i n l e t systems greatly reduce sample throughput and ov e r a l l system r e l i a b i l i t y . The step c o n t r o l l i n g the rate at which 10 analyses can be performed i s often the:time required for the replacement of carbon dioxide samples within the i n l e t system. A dual sided gas i n l e t system has been constructed out of s t a i n l e s s steel that features e l e c t r i c solenoid valves that are i n t e l l i g e n t l y controlled by a Motorola 6802 microprocessor.. At optimum speed a s a t i s f a c t o r i l y analyzed aliquot of carbon dioxide can be replaced by a new sample, pressures equalized and data c o l l e c t i o n i n i t i a t e d i n l e s s than 45 seconds.. The software features f a i l p r o o f design making i t very improbable that an operator can accidently i n i t i a t e an undesirable sequence. A f i n a l problem related s p e c i f i c a l l y to the automation of the: University of B r i t i s h Columbia system i s i n the area of data reduction. The mass spectrometry laboratory at the University of B r i t i s h Columbia operates two Interdata computers, one of which has been used to acquire and d i g i t a l l y f i l t e r data from the oxygen mass spectrometer. . Previously more sophisticated data reduction was done o f f - l i n e using the f a c i l i t i e s of the University of B r i t i s h Columbia's Computing Centre. The system was workable but not at a l l designed to maximize the number of samples that could be analyzed d a i l y . the automated system designed for t h i s t h e s i s 11 project the Interdata computer program has been modified so that i t serves both as the c o n t r o l l e r , i n i t i a t i n g various actions that the i n t e l l i g e n t devices on the mass spectrometer can perform, and also as a more sophisticated means to reduce the data. , The computer program monitors the standard error of the mean of the measured DEL values and can make decisions as to whether a new sample should be analyzed, the present sample should be reanalyzed or whether the analysis should just continue.. The system as a whole i s a good example of distributed processing.. The mass spectrometer system described i n t h i s t h e s i s has been designed i n such a manner that an automated sample preparation l i n e can be added at a l a t e r date. , The design of t h i s l i n e i s f a i r l y complete:but w i l l not be discussed i n this t h e s i s . 12 IX. IMPBOVEMENTS IN TBADITIONAL MASS SPECTEOMETBY 2.1 A New Technique of Gas Source Construction Design and i n s t a l l a t i o n of mass spectrometer ion sources have t r a d i t i o n a l l y been empirical i n nature as pointed out by Barnard (1953). An ion source for a gas source.mass spectrometer generally consists of four parts: 1. .. a gas i n j e c t i o n system 2. an io n i z a t i o n chamber where gas molecules traverse an electron beam 3. , an electron gun assembly 4. _ an ion gun designed to produce a well collimated beam of ions A well designed ion source should possess high s e n s i t i v i t y and also provide reasonable ion o p t i c a l properties.. S e n s i t i v i t y i n a gas source mass spectrometer can be defined as the ratio of the number of ions reaching the detector to the number of gas molecules passing through the i n l e t leak. . An electron gun assembly and io n i z a t i o n chamber were constructed and used in conjunction with a thick lens ion o p t i c a l system described by Loveless and Bussell (1969). These components were sucessfully i n s t a l l e d as a complete ion source in the.oxygen isotope mass spectrometer. . Commercially manufactured mass spectrometers are quoted 13 as having s e n s i t i v i t i e s of approximately 3.5 x 109 ions/s for a gas flow rate of 1.1 x 10 1 S molecules/s y i e l d i n g a s e n s i t i v i t y of 3.2 x 10 - 6 ions/molecule "or about one ion per 300,00 0 molecules. By taking advantage of the extremely large dispersion that a 30 cm, 90 degree sector f i e l d mass spectrometer possesses i n the mass range of i n t e r e s t , several source modifications allowed the s e n s i t i v i t y to be increased by about a factor of 50 over the previous mass spectrometer's.. The construction of the gas source used a new material made by Corning and named MACOR-Machinable Glass Ceramic. MACOR i s a material that has a unique p a r t i a l l y - c r y s t a l l i n e structure which enables i t to be machined to precise tolerances with ordinary metal working tools.. I t i s extremely non-porous and a very good e l e c t r i c a l i n s u l a t o r but can be coated with a layer of s i l v e r (Dupont #7713 conductive s i l v e r composition) to make any portion of i t e l e c t r i c a l l y conductive. This material has proven to be an excellent choice f o r the. source construction because i t allows adjustments to be made e a s i l y during source assembly and l a t e r when changing filaments. Figure 2. 1 sho.ws a cross section of the gas source assembly i n the XY plane.. (It should be noted at t h i s time that i n mass spectrometry the X axis i s in the di r e c t i o n of the emerging ion beam, the Y axis i s the narrow dimension of the beam and the Z axis i s the dire c t i o n of the magnetic field) . The' Macor portion of the source: follows the FIGURE 2.1 SIDE VIEW OF MACOR GAS SOURCE This cross-section of the: machinable glass gas source shows the r e l a t i v e positions of the various electrodes within the:gas source. A l l electrodes are machined from s t a i n l e s s s t e e l . The portions of the ionization chamber shown with a dashed l i n e have been coated with conductive: s i l v e r composition to prevent charge buildup.. FILAMENT POSTS I N C H A D J U S T M E N T SCREW ANODE ADJUSTMENT SCREW x q — U — f FILAMENT-CARBON DIOXIDE ENTRANCE R E P E L L E R T R A P SLIT ELECTRON BEAM DEFINING SLIT IONIZATION CHAMBER MACOR SOURCE CONDUCTIVE SILVER ION BEAM EX IT HOLE 16 conventional design of Nier (1947).. The filament i s spot welded onto filament posts that are e l e c t r i c a l l y insulated from the filament block.. The filament block i s equipped with a screw adjustment mechanism which allows the filament to be precisely located behind the exit s l i t . The filament used i s thoriated tungsten which has proven to be extremely long l i v e d . , We have obtained 2 1/2 years of nearly continuous service from one such filament as opposed to the previous ion source : which required filament replacement about every 6 months.. The anode and electron drawout electrodes produce a f i e l d which exerts a force on the electrons causing them to enter the the i o n i z a t i o n chamber with an energy of approximately 100 eV.. They are moderately collimated by the s l i t i n the electron drawout*. The carbon dioxide enters the i o n i z a t i o n chamber through the hole i n the:repeller electrode. The electron beam traversing the i o n i z a t i o n chamber travels i n a h e l i c a l path due to the presence of a magnetic f i e l d whose d i r e c t i o n i s p a r a l l e l to the electron beam. The r e s u l t i n g trapped electrons c o l l i d e with the molecules and some of these c o l l i s i o n s r e s u l t i n the creation of ions. Host of the electron beam passes through the i o n i z a t i o n chamber unperturbed and i s collected in the trap. The ions that are created are accelerated towards the c i r c u l a r e x i t of the source due to a positive p o t e n t i a l placed on the r e p e l l e r . . The emerging ion beam then enters 17 the Loveless source. I t i s worth mentioning that the: current through the filament i s regulated by the t o t a l number of electrons being emitted by the filament* This electron current i s t y p i c a l l y 200 microamperes. , Many of the electrons s t r i k e the.edge of the i o n i z a t i o n chamber and for t h i s reason a l l walls of the ionization chamber have been coated with conductive s i l v e r . Electrodes are generally held i n place with set screws with the exception of the r e p e l l e r which screws in t o threads i n the MA COB. The entire assembly f i t s on top of a modified Loveless source shown i n Figure 2.2 (Loveless and Bussell, 1969). This source was designed by Loveless to match source emittance and mass spectrometer admittance thus maximizing the number of ions which are produced in the i o n i z a t i o n chamber and leave the exit s l i t . The thick lens assembly works well with the newly designed Macor portion of the gas source. Minor modifications to the high voltage:supply were necessary to provide potentials appropriate to the various electrodes i n the Loveless source.. S e n s i t i v i t y of the arrangement was so great that the use of a source magnet might even be unnecessary.. However i t was decided that lower operating pressures and resulting increase in filament l i f e outweighed the advantages of omitting the source magnet. 18 FIGURE 2.2 THE COMPLETE GAS SOURCE Side view of the Macor gas source assembly mounted on top of the Loveless type thick lens assembly.. The f i v e s t a i n l e s s s t e e l plates are separated by ceramic b a l l bearings that insure proper alignment of source s l i t s * . The entire assembly mounts on the beam tube baseplate using just four bolts. I I N C H 1 20 2.2 The Dual Beam Ion C o l l e c t o r Multiple beam c o l l e c t i o n i s an inherently more: stable method than single beam c o l l e c t i o n since the quantity being measured i s the r a t i o of two or more ion beams.. It i s easy to see that variations i n source e f f i c i e n c y in time are minimized since they affect a l l peaks equally*„ To c o r r e c t l y c o l l e c t two ion beams simultaneously, consideration of the ion o p t i c a l properties of mass spectrometers i s extremely important since i t i s not at a l l clear where the defining s l i t s i n the c o l l e c t o r should be placed. although Dempster u t i l i z e d the focussing action of a 180° magnetic sector f i e l d as early as 1918, i t was l e f t to Herzog (1934) to publish a general analysis of the focussing action of magnetic f i e l d s . The Herzog equation places a constraint upon the : geometrical configuration a mass spectrometer may possess and s t i l l exhibit f i r s t order d i r e c t i o n focussing* , Stated another way, the Herzog equation indicates where the f o c a l point i s for a given source:and analyzer geometry. The Herzog equation i s : r 8 i n ( : « ( , ) + L x C O s ( ( } ' - e l ) + £ /cos (cj>-e 2) _ L ^ i n C ^ - e ^ ) ^ [ ^ } cos(e^) 2 \;cos(e2) rcos (e^) cos ^2) J where r i s the radius of curvature <j) i s the angle formed by the normals to the ion beam at the entrance to and exit from the magnetic f i e l d L1 i s the distance the ion beam travels from the source 21 e x i t s l i t to the entrance to the magnetic f i e l d e1 i s the angle the ion beam makes with the normal to the magnetic f i e l d at the entrance point L2 i s the distance the ion beam travels from the e x i t point to the f o c a l point e2 i s the angle the ion beam makes with the normal to the magnetic f i e l d at the e x i t point The f o c a l point determined by the Herzog equation i s defined to be the point where an ion beam, that i s diverging by a small angle when i t leaves the source, w i l l have i t s outer diverging rays and the: central ion beams brought closest together. The Herzog equation i s only f i r s t order in the sense that the incident angle of the:three rays i s assumed to be small (i.e..the rays are nearly p a r a l l e l ) , and i t i s therefore only valid for ion beams with small angles of divergence. Stacey (1962) previously investigated the focussing properties of a symmetrical sector f i e l d analyzer i n t h i s laboratory.. In his study, Stacey used Cartan's (1937) geometrical construction. The method proposed by Herzog was compared with Cartan's method using a Fortran IV program. This study showed that for symmetric configurations of source, c o l l e c t o r and analyzer (L1=L2 and e1=e2=0 i n equation 2 . 1 ) , the two methods y i e l d the same foc a l point. However, as the geometry of the system i s altered, the.two methods give s l i g h t l y d i f f e r e n t solutions and i n extremely asymmetric positions the f o c a l points calculated by the two 22 methods can d i f f e r by centimeters. , Table 2.1 summarizes the results of a single source analyzer geometry for a r e l a t i v e l y asymmetric configuration. 23 MASS |RADIOS | HERZOG | CARTAN | E2 | L2 -+ -+ •+ # i (cm) J X (cm) 1 Y (cm) |X(cm) 1 Y (cm) 1 ° 1 (cm) 40 | 28. 74 |79.95 1 -2. 10 |79.02 1-3.31 1 -7. 58 | 22.7 2 4 1 |29.09 i 80.98 1-1.95 |80.20 1 -2. 95 1 -6. 78 |23.52 42 |29.45 I 82.02 1-1.79 181.37 | -2.6 0 1 -6. 00 |24.34 43 | 29. 80 |83.07 1-1.63 182.54 1 -2.26 i-5. 26 125. 17 44 |30.14 |84. 12 1-1.45 183.7 1 1-1.94 1 -4. 54 |26.0 1 45 130.48 |85. 18 1-1.27 |84.87 1-1.63 1 "3. 85 126.87 46 130.82 |86.25 1-1.09 |86.03 1-1.34 1 -3. .18 1 27.74 47 131.15 |87.33 1-0.89 187. 18 1-1.06 1 -2. 53 128.62 48 |31.48 188.41 1-0.69 188.32 1-0.79 1 -1. 91 |29.52 49 131.81 I 89.51 1-0.49 |89.46 1-0.53 1 -1. 30 130.43 50 |32. 13 i90.61 1-0. 28 |90.60 1-0.29 \ -0. 71 |31. 35 -I TABLE 2.1 COMPARISON OF FOCAL POINTS DETERMINED BY HERZOG AND CARTAN METHODS Results obtained assuming the source e x i t s l i t was located at (0,0), apex of magnet at (46* 00,0.0), L 1=32.53 and the ion beam enters the f i e l d at a rig h t angle* Note: symmetric position of magnet i s (43.10,0.0).. Close examination of the r e s u l t s obtained from the two methods tends to support the use of Herzog's equation. 24 Figure: 2.3 i s a plot of the f o c a l points determined by the two methods. As can be c l e a r l y seen the Herzog focal points l i e on a r e l a t i v e l y straight l i n e whereas Cartan's construction yields f o c a l points that l i e on a c i r c u l a r arc of large radius (about 120 cm). Since Cartan's method approximates the boundary of the magnetic f i e l d by a c i r c u l a r arc (in t h i s case of 30 cm radius) and Herzog's method does not make t h i s approximation, i t i s believed that the Herzog method i s more r e l i a b l e . Since two ion beams are coll e c t e d simultaneously i n our mass spectrometer, an understanding of the focussing properties of a sector f i e l d magnet i s necessary. In such a case, i t i s esse n t i a l to place the defining s l i t s as close to the appropriate focal points as possible. In the previous oxygen isotope f a c i l i t y l i t t l e attention had been paid to the focussing of the mass 44 ion beam (mass 46 was in focus) and as a resu l t the defining s l i t was out of focus by approximately 2.4 cm* This d e t a i l i s often overlooked by other oxygen isotope f a c i l i t i e s . . The locations of the mass 44 and mass 46 foc a l points were determined for a symmetrical configuration of source and magnet with the assumption that the mass 45 central ion ray would have a radius of curvature of 30.48 cm (12 inches). The results indicated that the.mass 44 focal point was located within the c o l l e c t o r mounting assembly.. I t was therefore impossible to place the mass 44 defining s l i t at the correct position.. Since i t i s extremely important to FIGURE 2.3 A COMPARISON OF TWO METHODS OF FOCAL POINT DETERMINATION The positions of f i r s t order f o c a l points determined by the Herzog (circles) and Cartan (triangles) methods, for the asymmetric configuration of source, analyzer and c o l l e c t o r used in t h i s project. The positions for mass to charge r a t i o s from 40 to 50 are plotted.. The mass 4 5 ion beam was the central mass.. CM en SOURCE E X I T S L I T (0.0,0.0) M A G N E T A P E X (42.0,0.0) HORIZONTAL (CM) 27 keep mass 45 near the center of the beam tube the decision was made to use an asymmetrical configuration of source, analyzer and c o l l e c t o r . This i s a non-typical but c e r t a i n l y acceptable solution. Three d i s t i n c t motions of the magnet assembly are possible: movement in the d i r e c t i o n of the source-collector l i n e , movement in the dir e c t i o n perpendicular to t h i s l i n e , and a rotation of the magnet. These three, movements were studied t h e o r e t i c a l l y and i t was found that movement of the magnet along the source-collector l i n e had the greatest effect on the position of the f o c a l points. Figure 2.4 i l l u s t r a t e s the res u l t s of moving the ; magnet along the source ^ c o l l e c t o r line;. In a l l cases shown, the source i s assumed to be at the or i g i n and the ion beam enters the magnetic f i e l d perpendicular to the f i e l d boundary. as apparent from the figure there i s not a unique magnet location which gives satisfactory f o c a l point positions*, For t h i s project the solution obtained by moving the magnet 1.10 cm ho r i z o n t a l l y towards the source was chosen.. Table 2.2 summarizes the t h e o r e t i c a l data from which the dual c o l l e c t o r was designed* FIGURE 2.4 THE EFFECT OBTAINED BY MOVING THE MAGNET IN A HORIZONTAL DIRECTION The positions of the mass 44 ( c i r c l e s ) , mass 45 (triangles) and mass 46 (squares) f o c a l points for di f f e r e n t horizontal positions of the:magnetic analyzer. The positions of the apex of the magnet are indicated within brackets immediately to the l e f t of each l i n e . I t i s clear that the physical construction of the c o l l e c t o r mounting assembly made i t impossible to use a symmetric analyzer configuration for the dual beam c o l l e c t o r . . 0 MASS 44 A MASS 45 • MASS 4 6 2 . 0 i i i } i 8 5 . 0 8 6 . 0 6 7 0 8 8 . 0 8 9 . 0 HORIZONTAL POSITION (CM) N3 30 M A S S J-— RaDIOS FOCaL POINT # 1 IHorz.., I Vert. I I | . cm 1 cm | cm I ° I cm 43 | 29.80 I 84.34 I 0.07 I o. 18 | 29.99 44 | 30.14 | 85.58 I 0. 32 I o. 84 | 31.05 45 | 30.48 | 86.84 I 0. 58 I 1-47 i | 32.13 46 | 30.. 82 | 88.11, | 0.85 I 2. 08 j 33.23 47 | 31.16 | 89.40 I 1.13 1 2. 67 | 34.35 E2 T 1 L2 Coordinates of source exit s l i t (0.00,0.00) apex of magnet=(42.00,0. 00) at 90° L1=29.70 cm E1=0.00 Coordinates of point where beam i n i t i a l l y i n t e r s e c t s magnet field= (21.00,'21.00) TABLE 2.2 FOCAL POINT LOCATIONS FOR A SPECIFIC ASYMMETRIC CONFIGURATION OF SOURCE AND MAGNET Figure 2.5 shows the c o l l e c t o r assembly that was designed for the dual beam mass spectrometer. The f i r s t three s l i t s the ion beam encounters are the.mass 44 defining s l i t , a grounded supressor s l i t , and the secondary electron s l i t held at -45 V., The next three s l i t s are the equivalent s l i t s for mass 46. The addition of the v e r t i c a l suppressors insures that any ions or secondary electrons associated with FIGOBE 2.5 THE DUAL BEAM COLLECTOB ASSEMBLY This side view of the dual beam c o l l e c t o r assembly shows the r e l a t i v e positions of the various electrodes within the c o l l e c t o r . . They dimensions of the various s l i t s are indicated in the diagram. The v e r t i c a l suppressors are shown attached to the two mass 46 suppressor plates. . The superior method of heat sinking the parametric amplifiers i s also shown* BEAM TUBE BASEPLATE one ion beam do not af f e c t the other. . Since the i n s t a l l a t i o n of th i s c o l l e c t o r we have, experienced no d i f f i c u l t i e s associated with secondary electrons. F i n a l adjustments i n focussing are made empirically by mapping the ion beam peak shape for diff e r e n t locations of the magnetic analyzer* . Peak shape was determined by magnetically scanning over the mass 46 ion beam. The.ideal peak shape i s a trapezoid with a peak top of Wc-Wi and a peak bottom of Wc+Wi. . Wc i s the width of the c o l l e c t o r defining s l i t and Wi i s the image width which i s the sum of the source e x i t s l i t width and the spread of the ion beam given by Stephens (1934). 2 w. = w + l s ra ,2 , 2 • 7 , 2 L„ + r L. + r + T- 2 4- 2 , ;t 2 , 2 • I- 2 + r [2.2 ] where r i s the radius of curvature for mass 46, L1 and L2 are as described f o r equation 2.1 and a i s the:half angle of diverence of the ion beam e x i t t i n g the source.. Using the data i n Table 2.2 and noting e2 i s small so that Stephens' formula i n 2*2 applies we get Wi = Ws + 0 . 9 7 5 r a 2 = Ws + 30.05ct2 [2 .3 ] in t h i s mass spectrometer ffs = 0.008 inches, Wc(46) = 0.070 inches and Wc(44) = 0.140 inches, yielding a t h e o r e t i c a l best peak shape of peak shape(46)= (Wc-Wi)/(Wc+Wi) =0. 75 [2,4a] 34 peak shape(44) = (Wc-wi)/(Wc+Wi)=0.87 [2.4b] and r e a l i s t i c shapes of (a=0.5°) peak shape(46)=0.73 and peak shape(44) =0. 85. The magnetic analyzer was focussed empirically by f i r s t lowering i t s apex one pole gap width to compensate for f r i n g i n g f i e l d s * (Nier, 1940) and then systematically moving the magnet u n t i l peak shape (defined in 2.4a and 2.4b) and symmetry were maximized... In practice the large c o l l e c t o r s l i t widths used make t h i s method moderately i n s e n s i t i v e to magnet position but peak shapes obtained were extremely good with peak shape (46) =0.72 [2.6a] peak shape(44) =0. 86 [2.6b] The r e s u l t s given above validate the theoretical work from which the dual beam c o l l e c t o r was designed* , The improved flatness of the peak tops of in t e r e s t i s a d i r e c t result of the care taken in the placement of the defining s l i t s and the suppression of any secondary electrons that are created i n the c o l l e c t o r . . As in the previous f a c i l i t y described by Russell and Ahern (1974), the preamplifiers have been mounted inside the vacuum area. However, the new design provides a much more r i g i d mounting system as well as a means of heat sinking the amplifiers.. D i s c o n t i n u i t i e s in the baselines, previously 35 observed with these amplifiers, have been eliminated p a r t l y by the superior heat sinking and partly through improved s t a b i l i t y of the vacuum inside the mass spectrometer. 2.3 Direct Regulation of the Magnetic F i e l d The radius of curvature of an ion beam i n a mass spectrometer i s determined by r=(2Vm/g) V 2/B [2.7] where m/g i s the mass to charge r a t i o of an ion beam, V i s the accelerating voltage and B i s the magnetic f i e l d i n t e n s i t y . . Neglecting variations i n source beam emittance, the s t a b i l i t y of an ion beam of a given mass to charge ra t i o , i s determined s o l e l y by the s t a b i l i t y of the accelerating voltage and the magnetic f i e l d i n t e n s i t y . , I t follows from 2.7 that dr/r=-dB/B+dV/2V [2.8] Examination of equation 2.8 indicates that the s t a b i l i t y of the position of the.ion beam i s more dependent upon the s t a b i l i t y of the magnetic f i e l d i n t e n s i t y , B, than on s t a b i l i t y of the high voltage.. It was also necessary to implement computer control of the radius of curvature r and there were p r a c t i c a l d i f f i c u l t i e s related to a l t e r i n g the high voltage, V, applied to a Loveless source. For these reasons i t was decided to improve the regulation of the magnetic f i e l d * 36 In mass spectrometry laboratories i t i s not uncommon to regulate the magnetic f i e l d i n d i r e c t l y by regulating the current through the magnet c o i l s . . This approach has d i f f i c u l t i e s because hysteresis destroys the simple functional relationship between magnetic f i e l d i n tensity and the current passing through the magnet d o i l s . . f. The a v a i l a b i l i t y of commercial gaussmeters make d i r e c t regulation of the magnetic f i e l d a f a i r l y simple task.. A gaussmeter (Bell model 640) was obtained by t h i s laboratory and i n s t a l l e d i n the new oxygen isotope mass spectrometer. Figure.2*6 i s a block diagram showing how the gaussmeter was i n s t a l l e d in the system.. The gaussmeter operates in an incremental mode where i t s output i s determined as follows V(Bell)=G(B-K) [2.9] where G i s the gain determined by a s e n s i t i v i t y control on the:front panel, B i s the magnetic f i e l d i n tensity and K i s a constant which i s e f f e c t i v e l y subtracted from the t o t a l f i e l d . „ The output of a motor driven scan pot i s buffered by a non-inverting operational amplifier and i s added into the feedback loop as i s the output of a d i g i t a l magnet co n t r o l l e r (see next chapter) . The output of the summing amplifier i s fed into a Kepco Model OPS-40 Power Supply which i s functionally equivalent to an extremely high power operational amplifier. The output of the Kepco provides a current through the 22K r e s i s t o r into the n u l l i n g input of the trigger board FIGURE 2.6 CIRCUITRY USED FOR DIRECT REGULATION OF MAGNETIC FIELD The i n c l u s i o n of the B e l l Model 640 gaussmeter in the feedback loop of the magnet current supply i s shown i n t h i s f i g u r e . . The output of the D i g i t a l Magnet Controller i s summed along with the output of a motor driven scan pot by the 74 1 summing amplifier. The Jepco power amplifier provides the co n t r o l s i g n a l to the t r i a c controlled A.C. power supply, as well as providing high frequency c o n t r o l of the magnetic f i e l d strength.. TRIAC CONTROLLED D.C. V O L T A G E SOURCE + 15 - V W M A G N E T C O I L POLE P I E C E S H A L L ( J ) P R O B E -vW-h - V V \ A - 4 - V v > A O V E R C U R P E N T P R O T E C T I O N CIRCUITRY SUMMING AMPLIFIER KEPCO MODEL OPS 40 0.5(C) POWER AMPLIFIER DIGITAL MAGNET I CONTROLLER <R2 WV-Rl • R3 M A G N E T COIL CURRENT C M METER BELL MODEL 6 4 0 GAUSSMETER 1 B U F F E R M A N U A L S C A N POT - 6 —-VvV— + 4> LO oo 39 described by Russell and B e l l i s (1971) and also provides high frequency, low amplitude regulation at the output of the Triac controlled power supply. The magnetic f i e l d i n t e n s i t y generated by the magnet c o i l s i s l i n e a r l y related to the output of the; Kepco power supply.. This output i s a function of 1.. the output of t h e i B a l l gaussmeter, V(Bell) 2. the voltage input from the d i g i t a l magnet c o n t r o l l e r , V(DMC) 3.. the voltage input from the buffered motor driven scan pot, V (SCAN) . .. In f a c t B«R4oV (Bell) /R1+R4«V(DHC) /R2+R4«V (SCAN) /R3 [ 2. 10 ] I t can be seen that the magnetic f i e l d i s a l i n e a r function of the control voltages. None of the control elements a f f e c t the closed loop gain of the c i r c u i t , a statement which was not true.for the c i r c u i t described by Russell and B e l l i s (1971).. K o l l a r (1960) indicated that a previous magnet co n t r o l l e r produced a current that was stable to 1 part i n 25,000 for periods of seconds or minutes. The: c i r c u i t described by Russell and B e l l i s (1971) , was found to regulate the magnetic f i e l d to 1 part i n about 16,000. After the c i r c u i t of Figure 2.6 was i n s t a l l e d the magnetic f i e l d was found to be regulated to 1 part i n 54,000, an improvement of more than a factor of 3. 40 The Hall probe used with the B e l l gaussmeter i s temperature compensated and the long term d r i f t i s small enough to permit unattended sample analysis for up to 6 hours.. Figure 2 .7 shows the res u l t s of analyzing a working standard for a period of 8 hours. The sharp discontinuity in the measured values after 6 hours was d i r e c t l y traceable to d r i f t i n either the magnet current supply or in the high voltage supply. At the conclusion of t h i s experiment i t was clear that no major design flaws existed i n the instrumentation and longer term s t a b i l i t y could only be obtained by more sophisticated techniques of d r i f t compensation (see Chapter V). FIGURE 2.7 THE STABILITY OF MEASURED DEL VALUES USING DIRECT FIELD REGULATION The results of analyzing i d e n t i c a l samples of carbon dioxide i n both sides of the i n l e t system are shown i n the.following figure.. The:DEL values obtained are indicated by the upper l i n e (circles) and the calculated error i n the DEL measurements i s shown in the lower l i n e ( t r i a n g l e s ) . . The s t a b i l i t y of the.mass spectrometer was limited to about 320 minutes. The v a l i d i t y of the method of determining the error i n calculated DEL values i s also c l e a r l y shown by t h i s figure. . I I I . INNOVATIONS IN DIGITALLY CONTROLLED MASS SPECTROMETRY 3.1 A Brief Description of Peripheral Interfacing to Interdata Minicomputers Since several of the sections i n t h i s chapter assume some knowledge of the techniques Interdata computers use to communicate with peripheral devices, i t seems appropriate to present a b r i e f summary at the outset of t h i s chapter. More complete information may be: obtained by consulting D.L..Mitchell (1971) or the appropriate Interdata manuals. None of the c i r c u i t r y described i n t h i s thesis causes interrupts of the Interdata computer and therefore no discussion of interrupt processing i s presented.. The multiplexor bus of the Interdata minicomputer consists of 27 unidirectional l i n e s which can be divided into four basic groups: '1._ Eight data available l i n e s (DALS) that carry information from the minicomputer to-the perpheral device. 2., Eight data ready l i n e s (DRLs) that carry information from the peripheral device to the minicomputer.. 3. „. Eight control l i n e s that control the use and intent of the DALs and DRLs. These include data available (DA) and command (CMD), which indicate:when there i s v a l i d data on the DALs; status ready (SR) and data 44 ready (DR) which indicate when the data on the DHLs i s clocked into the minicomputer; the address command l i n e (ADRS) which indicates when an address i s present on the DALs; and two interrupt control l i n e s . 4.. Three miscellaneous l i n e s including system synchronization and system cle a r l i n e s . A t y p i c a l operation might be as follows: the processor puts the 8 b i t address of the external device on the DALS; the ADRS l i n e i s raised; the external device with that address sets a f l a g which enables i t to respond to subsequent control l i n e t r a n s i t i o n s ; the processor places data on the DALs; the processor raises either CHD or DA (depending on the i n s t r u c t i o n being executed).. The peripheral device t y p i c a l l y uses CMD or DA to clock the information on the DALs into the appropriate hardware within the device. , Input operations are quite similar to the: above with the exception that the external device i s the source of the data on the DRLs and the.processor raises either the DR or SR command l i n e s to acquire the information.. P.Hhaite was responsible for extending the multiplexor bus to the mass spectrometers using o p t i c a l i s o l a t i o n techniques.. The i n d i v i d u a l l i n e s of the bus are available at each external device in several forms including: buffered, gated, active low, active high, pulsed high and pulsed low. An example of the terminology used to describe 45 which l i n e i s being referred to i s DAG1P which means the o p t i c a l l y i s o l a t e d data available l i n e , gated and pulsed temporily to a high (1) state., H. Verwoerd was responsible for the design of the in t e r f a c i n g which made t h i s multitude of multiplexor l i n e states available.. These, c o n t r i b u i t i o n s greatly reduced the task of c i r c u i t r y design.. 3.2 The D i g i t a l Magnet Controller The long term s t a b i l i t y needed i n thi s mass spectrometer required the inclusion of d i g i t a l c i r c u i t r y that could i n t e r a c t simply with a more powerful c o n t r o l l e r , the Interdata minicomputer. As was mentioned i n the previous chapter, s t a b i l i t y of the high voltage, and/or magnetic f i e l d was limited to about 7 hours.. The objective of t h i s t h e s i s project was to design a mass spectrometer system that would have.long term s t a b i l i t y approaching 24 hours. .. The above observation prompted the development of the more sophisticated approach. . The approach taken was to design a d i g i t a l magnet co n t r o l l e r whose output could a l t e r the strength of the mass spectrometer's magnetic f i e l d . The reasons f o r c o n t r o l l i n g the magnetic f i e l d instead of the.high voltage:were twofold; f i r s t l y , i t was much easier to add a control voltage to the feedback loop of the magnetic f i e l d ; and secondly, an unclean Loveless source.has an extremely long c h a r a c t e r i s t i c time when subjected to step changes i n the accelerating 46 potential (Ryan, personal communication, see section 3.6)., The design c r i t e r i a f o r the - d i g i t a l magnet c o n t r o l l e r were as follows: 1. . The c o n t r o l l e r had to possess both extremely high resolution (many points over a peak) and at the same time have a large dynamic range (the a b i l i t y to scan a large portion of the spectrum). 2. _ It should respond to increments i n the magnetic f i e l d instead of absolute f i e l d values* This w i l l eliminate the need for the minicomputer to know i n i t i a l settings of the c o n t r o l l e r . . 3. ,„ It should be possible to increment or decrement the magnetic f i e l d . 4. „ I t should be able to automatically scan over peaks without interaction from the minicomputer,. 5. . It should be able to reset the c o n t r o l l e r to a mid-range position either manually, under computer i n i t i a l i z a t i o n or under program c o n t r o l . 6. . I t should sense overflow or underflow conditions so that the minicomputer can take appropriate action* , Figure 3.1 i s the l o g i c diagram for the d i g i t a l magnet con t r o l l e r , k b r i e f description of the. c i r c u i t follows* The 3 bit control r e g i s t e r i s loaded with a control byte using an Output Command to the d i g i t a l magnet c o n t r o l l e r . Depending on which b i t (s) are set, the c o n t r o l l e r may do one of eight things: 1. . enable the peak side most s i g n i f i c a n t bits ( M S B ) FIGURE 3.1 DIGITAL MAGNET CONTROLLER CIRCUITRY The following figure i s a block diagram showing the operation of the d i g i t a l magnet c o n t r o l l e r . The multiplexor bus shown i n the l e f t of the figure i s the o p t i c a l l y i s o l a t e d bus present on each interface card.. The s t a b i l i t y of t h e . d i g i t a l feedback loop i s insured due to the delay inherent within the d i g i t a l adder c i r c u i t s . . A CLIAH I • MAN OFF t EIGHT BIT CONTROl. REGISTER 2HZ CLOCK ENABLE ENABLE SCAN MSB DATA LATCH SCAN LSB DATA LATCH PEAK MSB DATA LATCH PEAK LSB DATA LATCH — 1 = ENABLE SCAN MIWANH i — i f r I LSB I [MSB I I LSB I SCAN 10 BIT BINARY ADOER SCAN OVERFLOW y y PEAK OVERFLOW PEAK 10 BIT BINARY AOOER MIORANGE CIRCUITRY DIGITAL FEEDBACK 1 SCAN 10 BIT D/A LATCH 21 3 > 0 PEAK 10 BIT D/A LATCH ITCH 10 LED DRIVER 10 BIT 0/A n 9^ 10 BIT D/A 10 LEO DRIVERS 49 data la t c h 2. _ enable the peak side less s i g n i f i c a n t b i t s (LSB) data latch 3. „. enable the scan side most s i g n i f i c a n t b i t data l a t c h 4.. enable the scan side least s i g n i f i c a n t b i t data l a t c h 5. _ clock the peak side D/A latch 6. a clock the scan side D/A latch •7. _ enable the 2 Hz o s c i l l a t o r for automatic scanning using the scan side c i r c u i t r y 8. . reset the scan and peak sides to midrange. With the exception of the autoscanning c a p a b i l i t y on the scan side c i r c u i t r y and the fact that the output range of the scan side i s 18 times greater than the.output range of the peak side, the two sides of the d i g i t a l magnet c o n t r o l l e r function i d e n t i c a l l y * The ranges of the two sides are determined by the two input r e s i s t o r s and the feedback r e s i s t o r of the f i n a l operational amplifier.. The c o n t r o l l e r has a dynamic range and resolution bettering a 14 b i t D/A converter even though i t i s constructed from two much l e s s expensive 10 b i t converters.. The peak side c i r c u i t functions i n the following fashion. The peak side most s i g n i f i c a n t b i t data latch i s enabled by the appropriate output command, the 6 most s i g n i f i c a n t b i t s of a signed 10 b i t number ( f i e l d increment) are loaded into the most s i g n i f i c a n t b i t data latch using a Write Data i n s t r u c t i o n . The peak side least s i g n i f i c a n t b i t 50 data latch i s enabled and the 4 least s i g n i f i c a n t b i t s of the f i e l d increment are loaded i n t o the least s i g n i f i c a n t b i t data lat c h . The 10 b i t two's complement number stored in the peak data l a t c h i s input to a 10 b i t adder whose other input i s the value currently held i n the D/A l a t c h . The output of the 10 b i t adder w i l l equal the sum of the two inputs after a delay of about 150 ns, a time: that i s an order of magnitude less than the f a s t e s t instructions i n the Interdata minicomputer* At t h i s point t h e : f i e l d increment i n the data latches has appeared on the inputs to the D/A latch but has not yet affected the output analog voltage. By checking the carry b i t of the 10 b i t adder at t h i s time an overflow or underflow condition can be detected and the f i e l d increment/decrement cancelled or adjusted accordingly., I f the f i e l d was being incremented then the carry b i t being set indicates an overflow condition whereas i f the f i e l d was being decremented an underflow condition i s indicated by the carry b i t being reset. After sensing the status of the peak side carry b i t and determining that no over/underflow condition would r e s u l t , the control register i s altered so that the D/A latch i s clocked.. This tranfers t h e : f i e l d change requested to the inputs of the D/A converter, changing the f i e l d accordingly. It should be noted that the D/A latches used have a setup time of 10 ns and a minimum hold time of 0 ns which i n conjunction with the delay of 150 ns through the 10 b i t 51 adders insures the s t a b i l i t y of the d i g i t a l feedback loop. A device support routine (DSR) was written for the d i g i t a l magnet c o n t r o l l e r which greatly reduces the e f f o r t required by a programmer. This program was written i n Interdata Assembler language : and i s reproduced i n Appendix I. The c i r c u i t r y which resets the co n t r o l l e r to midrange (a l l b i t s reset except for the most s i g n i f i c a n t bit) i s quite complicated and so a b r i e f description i s appropriate* The reset pulse may originate either from a b i t in the control r e g i s t e r , the system clear control l i n e , or manual depression of a toggle switch. The manual switch i s debounced via the D f l i p - f l o p at the top center of the diagram and OR'd with the two other possible sources for the reset command. A l l latches i n the c i r c u i t are;reset by t h i s pulse and i n addition the most s i g n i f i c a n t b i t s of the inputs to both D/A latches are placed in a high state. The complementary output of a second D f l i p - f l o p i s input to an inverter with a capacitively loaded output, thus delaying the output pulse by a few tens of nanoseconds., This delayed pulse clocks both D/A latches after the reset pulse i s removed and thus sets the latches and D/A converters to midrange.. When the output actually goes high i t resets the second D f l i p - f l o p and the process i s completed.. The c i r c u i t was constructed using complementary metal oxide semiconducter (CMOS) integrated c i r c u i t s and has given no problems since i t was f i r s t i n s t a l l e d over one year ago. 52 The operational amplifiers used were inexpensive 741s since long term d r i f t c h a r a c t e r i s t i c s are unimportant due to a software algorithm that corrects for d r i f t s anywhere in the magnetic f i e l d or high voltage c i r c u i t r y . The regulation of the magnetic f i e l d was not affected by the : i n s t a l l a t i o n of t h i s device. 3.3 D i g i t a l Compensation of Preamplifier Baselines Russell and Ahern (1974) describe the parametric amplifier based measuring system used in t h i s mass spectrometer.. As mentioned i n that paper i t was evident that non-zero baselines could s i g n i f i c a n t l y a l t e r measured is o t o p i c r a t i o s . For t h i s reason the manual operation of the mass spectrometer always included a step where the outputs of the parametric amplifiers were nulled before performing an analysis.. It was clear that i f the automation of the f a c i l i t y was to succeed then c i r c u i t r y had to be designed that would allow the automatic zeroing of baselines. R.D. Russell and R.D.-Meldrum spent some time examining the f e a s i b i l i t y of using analog sample and hold amplifiers to sample the baselines of the amplifiers and eliminate them through a feedback network.. They concluded that no e x i s t i n g analog sample and hold amplifiers possessed low enough droop rates to allow t h e i r use. The method of successive approximation often used i n analog to d i g i t a l converters can 53 easily be modified to produce a d i g i t a l sample and hold c i r c u i t with an i n f i n i t e ' hold time*. The solution to the problem of automatic baseline zeroing therefore was found i n the realm of d i g i t a l , rather than analog, techniques.. The design requirements for the baseline: compensators were as follows: 1. . Baselines of the 1702 parametric amplifiers had to be within 5 mV of ground potential to eliminate errors of s i g n i f i c a n t magnitude in the measured isotopic r a t i o s (Ahern, 1975a)• , 2. _ The compensator should be able to n u l l baseline o f f s e t s of 0.150 V magnitude* 3.. The compensator should be activated by a simple negative going pulse*-4.. The compensator should indicate; that i t i s performing i t s task by setting a busy b i t that can be sensed by the Interdata minicomputer. Figure 3.2 i s a lo g i c diagram that describes the operation of the baseline compensators designed for t h i s system* The heart of the c i r c u i t i s an 8 b i t programmable counter (EXAR Integrated Systems, Inc.. Part number XH-224 0). The s p e c i f i c operation of t h i s chip can be understood by consulting the relevant data sheet. However, in the configuration used for the baseline compensation c i r c u i t i t i s esse n t i a l l y a counter with outputs which can be set to a high state by applying a high voltage to i t s reset input and w i l l then go low when an active high t r i g g e r FIGURE 3.2 DIGITAL BASELINE COMPENSATION CIRCUITRY A block diagram of the the d i g i t a l baseline c i r c u i t designed as an important portion of this thesis project* , Input and output signals are received from the Interdata minicomputer through the General Input/Output module described i n t h i s chapter. Offsets i n any analog components are compensated for with the exception of the Butterworth f i l t e r c i r c u i t since the signals used by the measuring system do not pass through the f i l t e r . The components within the dotted l i n e are not part of the baseline compensation c i r c u i t but instead are part of the: mass spectrometer measuring systems. Ol/'ANALOO COMPARATOR Ell 96L02 1£\ 1 96L02 B U » Y - i E N T E R H I G H V A C U U M d 1 Wv^ O R ix 1 0 ' n . [ — l / P A P A M E T R I C JL A M P L I F I E R | V (COMP) E X I T H I G H V A C U U M - 0 o • C 74LS7+ Ln 56 pulse i s detected. If the device i s enabled by the presence of an appropriate voltage at the disable input, then the device w i l l function as an ordinary binary counter. k brief description of the operation of the c i r c u i t follows. The three monostable vibrators at the bottom of Figure:3.2 are triggered by negative going pulses on the i r inputs. The f i r s t monostable vibrator w i l l apply an 80 microsecond reset pulse to the XB-2240. As the.reset pulse i s removed, the second monostable applies a trigger pulse.to the XB-2240. After 80 microseconds the trigger pulse i s removed and the thi r d monostable vibrator sets the D f l i p - f l o p connected to the disable input of the XR-2240, enabling the counter. The line:which enables the counter i s also the device's busy b i t which can be. externally monitored. ,. The outputs of the;counter are input to an 8 b i t D/A converter (manufactured by Precision Monolithics Incorporated as part number DAC-100) that i s connected i n a bipolar configuration. This converter serves as a current source for an operational amplifier.. I t should be noted that the feedback r e s i s t o r i s contained in the DAC-100 and so the output of the operational amplifier must swing between +5 V and -5 V.. The output of the operational amplifier passes through a second inverting amplifier with a gain of -1.. This output i s fed into a c i r c u i t quite s i m i l a r to that described by Bussell and Ahern (1974).. I t can be shown that the output of the parametric amplifier i s given 57 by: V=i[RfRo/Rc+RfRo/Rg +Rf]-VcRo/Rc+Vo [3.1] assuming the feedback r e s i s t o r Rf i s much greater than any other resistance and the battery switch i s open.. In the absence of an ion current i t i s clear that the output of the parametric amplifier i s V=Vo-VcRo/RC -0.135<V<+0.135 [3.2] where V i s the output voltage, Vo i s the. o f f s e t voltage and Vc i s the compensation voltage input from the c i r c i a t being described. Once the counter begins counting, Vc begins changing from -5 V to +5 V.. Sometime during t h i s stepwise t r a n s i t i o n Vc=VoRc/Ro [3.3] and the output voltage defined by equation 3.2 goes to zero which i s the.desired condition.. The output voltage passes through an i n v e r t i n g Butter worth f i l t e r with a 1 Hz cutoff frequency to eliminate 60 Hz noise.. I t then enters the non-inverting input of a precision comparator (PMI number CMP-02). . The inverting input of t h i s comparator i s t i e d to ground, as the output of the Butterworth f i l t e r passes through ground p o t e n t i a l , indicating a zeroed baseline for the 1702 parametric amplifier, the output of the comparator goes high, clocking the D f l i p - f l o p which disables the programmable counter and turns o f f the busy b i t . The voltage, gain of the comparator used i s t y p i c a l l y 58 500 V/mV which allows voltages to be compared to within a f r a c t i o n of a m i l l i v o l t . . The resolution of the 8 b i t A/D converter constructed i s one part i n 256 and since the dynamic range of VcRo/Rc i s -0.135 V to +0.135 V baselines can be zeroed to approximately 1 mV.: Offsets are unimportant i n a l l components except the Butterworth f i l t e r which was constructed with a PHI OP-07 ultralow o f f s e t amplifier. The most d i f f i c u l t problem encountered during the i n s t a l l a t i o n of these c i r c u i t s was related to noise on the outputs of the parametric amplifiers.. As much as 25 mV of 60 Hz noise exists on the u n f i l t e r e d baselines. When t r y i n g to n u l l baselines to within 5 mV with fas t d i g i t a l c i r c u i t r y , the presence of large amplitude noise i s a formidable problem. Normally the noise component i s removed by a d i g i t a l f i l t e r i n g scheme in the minicomputer.. D i g i t a l f i l t e r i n g was not desirable i n t h i s case but the contraction of a low pass Butterworth f i l t e r to perform analog f i l t e r i n g was straightforward. The only detrimental e f f e c t of such an action i s that the response.time increases as the cutoff frequency decreases. This necessitates running the programmable counter at a low rate of 50 Hz so that the result of a step change in Vc has time to pass through the c i r c u i t r y to the comparator.. For similar reasons, the delays b u i l t into the monostable vibrators must be long to allow for the slow response of the f i l t e r . . In practice t h i s means that the BUSY b i t should not be sampled for about 59 0.1 ms after issuing the s t a r t pulse to the: compensator. This i s e a s i l y achieved and does not warrant addi t i o n a l c i r c u i t r y to compute a more: involved busy b i t . . In practice the magnetic f i e l d i s altered by the con t r o l l e r described i n the previous section so that no ion beam enters the Faraday cups. Start pulses are given to two compensation c i r c u i t s , one f o r the mass 44 ion beam and another for the mass 46. A delay of approximately 0.5 s i s set and the minicomputer then st a r t s examining the busy b i t s for the baseline compensators u n t i l i t finds them both reset., At t h i s time baseline zeroing has been completed and the next step i n the analysis can begin. 3.4 The Design of a Multiplexed Analog to D i g i t a l Converter The discussion in Chapter V w i l l present many reasons why the. Interdata minicomputer must know the magnitude of analog voltages within the mass spectrometer.. In p a r t i c u l a r i t w i l l be shown that f a i r l y sophisticated algorithms such as automatic peak centering require information about the voltages present at the output of the parametric amplifiers in the measuring system. Other tasks such as measurement of i 2 c / » 3 c r a t i o s or **N/lsN ra t i o s are aided by the a b i l i t y to measure the peak heights of either ion beam d i r e c t l y . Although i t has not been included as a part of t h i s thesis project, the inc l u s i o n of t h i s 16 channel A/D converter has 60 made i t a simple task to monitor such things as magnet current, high voltage, source magnet current and beam tube pressure, to name just a few parameters. The recent a v a i l a b i l i t y of the:8700 series of CMOS A/D converters from Teledyne Semiconducter, made th e i r 8702 12 bi t converters an a t t r a c t i v e choice as the heart of t h i s c i r c u i t . . The converters are very inexpensive; ($25) and are s u f f i c i e n t l y f a s t (20 ms conversion time) f o r application i n the oxygen mass spectrometer., 16 channel analog multiplexors are now re a d i l y available i n the CMOS family of integrated c i r c u i t s . . This s i m p l i f i e d the design and construction of the multiplexed A/D converter described in the following paragraphs.. Figure 3.3 i s the logic diagram which shows the designed c i r c u i t . The c i r c u i t has been b u i l t on one of the o p t i c a l l y i s o l a t e d interface cards mentioned i n Section 3.2. A description of the c i r c u i t ' s operation follows.. The c i r c u i t begins functioning when an 8 b i t control word i s latched by CMDG1P into the two 4 b i t D type latches in the l e f t portion of the c i r c u i t . , CMDG1P w i l l also set the.JX f l i p - f l o p i n the bottom l e f t of the c i r c u i t . The upper latch i s fed into the selector inputs of a 16 channel multiplexor and one of the 16 analog voltages i s connected to the output of the multiplexor and i n turn to the D/A converter* The b i t pattern stored i n the lower latch determines which of four functions the c i r c u i t w i l l perform: 1.„ i n i t i a t e a conversion by enabling the output of the FIGURE 3.3 A 16 CHANNEL MULTIPLEXED ANALOG TO DIGITAL CONVERTER Up to 16 analog voltages can be sampled using the c i r c u i t r y shown i n the following figure.. The busy b i t must be sampled and found to be low before the data on the DINs i s r e l i a b l e * . The DINs are multiplexed and f i n a l l y appear on the DRLs whenever a Read Data in s t r u c t i o n i s processed for t h i s c i r c u i t * A VOLTAGE INPUTS (0-10 VI 4 BIT INPUT S E L E C T L A T C H 4 0 4 2 BIT CONTROL LATCH 40+3 VOLTAGE INPUTS (0-10 VI t(, CHANNEL A N A L O G MULTIPLEXOR 4.047 INITIATE CONVERSION SELECT a LSB'S SELECT «• MSB'S TELEDYNE SEMICONOUCTEH S702 A/0 CONVERTER •=L> FRONT 0.5 & PANEL 1 SHOT ' LEO SSS K3 63 upper monostable.vibrator 2. „ place the 8 l e a s t s i g n i f i c a n t b i t s on the DRLs 3. place the 4 most s i g n i f i c a n t b i t s on the 4 most s i g n i f i c a n t DRLs 4. , place the c i r c u i t in a configuration which w i l l f i r s t i n i t i a t e a conversion and then place the 8 le a s t s i g n i f i c a n t b i t s on the DRLs. . At the end of the f i r s t DEGOP (when the 8 least s i g n i f i c a n t b i t s are read) the JK f l i p - f l o p w i l l toggle and the 4 most s i g n i f i c a n t b i t s w i l l be placed on the DRLs. _ The appropriate busy b i t i s calculated by combining the BUSY and DATA VALID outputs of the 8702 converter. I t i s necessary to sense the status of the device.after issuing an i n i t i a t e conversion command and wait for a not busy indicati o n before sampling the d i g i t a l outputs. The monostable vibrator at the bottom of the c i r c u i t simply pulses a l i g h t emitting diode whenever a conversion i s i n i t i a t e d . . With the exception of the input latches which are LPSTTL, the device i s constructed using CMOS integrated c i r c u i t s . In practice the outputs of the address latch are translated to the 10 V l o g i c l e v e l s of the. 16 channel multiplexor by a 4104 low to high voltage translator.. The higher d i g i t a l l o g i c l e v e l allows analog voltages anywhere from 0 V to 10 V to pass through the multiplexor which means this device can only be used to measure positive voltages.. The resolution of the device i s 1 part i n 4096., Since i t s dynamic range i s 10 V t h i s y i e l d s a voltage resolution 64 of approximately 2.5 mV for a single conversion._ In the oxygen system, multiple sampling i s used to eliminate periodic noise and also to increase the resolution. In fact, i f the random noise i s of a magnitude greater than the voltage resolution of the converter, and there i s no systematic noise i n the. system, the resolution w i l l be increased by a square root of n factor* In t h i s system peaks are sampled 32 times, improving the resolution by a factor of approximately 5.6. This has the e f f e c t of giving th i s 12 b i t converter the a t t r i b u t e s of a 14 b i t converter. The voltage resolution i s improved to better than 0.5 mV by t h i s multiple sampling. 3.5 The General D i g i t a l Input/Output Interface Module Many of the. c i r c u i t s designed for the oxygen isotope f a c i l i t y possess only a few input control l i n e s and output status l i n e s . For this reason i t seemed inappropriate to dedicate a minicomputer address and interface board to each of these devices. . A general input/output (GIO) l o g i c module was designed that could issue l e v e l changes or pulses to any one of 16 outputs under program control.. R.D. Meldrum designed an input c i r c u i t that increased the number of input l i n e s that could be read or sensed from the o r i g i n a l l y planned 16, to 32.. This extension has not yet been u t i l i z e d . The design c r i t e r i a f o r the.GIO logic module were as follow s: 1. . The GIO logic module had to be able.to address 16 output ports. . 2... The output repertoire had to consist of active low 10 microsecond pulses, l e v e l t r a n s i t i o n s in either direction and the output module also had to be inactive when an input operation was requested. 3. . The GIO l o g i c module had to be able to either sense status on 8 status l i n e s or read data on 8 other data l i n e s . . The R.p Meldrum extension mentioned e a r l i e r also required the module to select either A side or B side inputs for both reading and sensing operations* figure.3.4 i s the logic diagram for the GIO module. The actual c i r c u i t was constructed using LPSTTL and CMOS integrated c i r c u i t s . . The input portion of the c i r c u i t i s l o g i c a l l y separate from the output c i r c u i t r y and i s shown i n the lower l e f t portion of the l o g i c diagram.. An output command issued to the device with DALO set w i l l connect the B side inputs to the two d i g i t a l multiplexors.. The f i r s t request for input from the GIO module w i l l cause the JK f l i p - f l o p to toggle and cause the A side inputs to again be selected. The outputs of the two d i g i t a l multiplexors are connected to the SINs or DINs and the information i s accessible to the Interdata minicomputer with either a sense status or read data i n s t r u c t i o n , respectively. . Normally, information should be accessed through the A side inputs. FIGURE 3.4 THE GENERAL INPUT/OUTPUT LOGIC MODULE The following f i g u r e - i s a l o g i c diagram for the GIO module. used to interface the Interdata minicomputer to many d i f f e r e n t components within the mass spectrometer system. The input c i r c u i t r y consists of the 3 integrated c i r c u i t s i n the extreme lower l e f t of the figu r e . . The 16 boxes at the top of the figure are the output l o g i c modules and the c i r c u i t r y i s as shown in the f a r l e f t module. INPUT 8 A AND 8 B 5INS 8-JINPUT MUX SINS CMDGIP i—I I OP lb DECODER VWLATtH DALS A 5 <• 1 T t t T INTERDATA MULTIPLE" *0R BUS A / B SIDE SELECT s>p\ Tow o» JaJci ft? K C J C 5 1 a oj 8 A AND B 8 DINS 8-? INPUT MUX DIMS O U T P U T LOGIC RESET PULSE ENABLE 5=> r-HI—KVVS MONO STABLE 10 US PULSE ON 68 The output portion of the GIO c i r c u i t functions in a very straightforward manner*. Dal4 through Dal7 are clocked into the input latch of the 16 channel multiplexor with CMDG1P.. After a delay generated by the monostable vibrator, the delayed CMDGOP clocks Dal2 and Dal3 into the two D latches. This delay i s necessary to insure that the output address i s selected before the appropriate gates are enabled. The c i r c u i t r y i n the lower r i g h t corner decodes Dal2 and Dal3 and determines whether a pulse or a high or low l e v e l change was requested. . Either the set or reset l i n e s w i l l go high or i n the case of a pulse;request, f i r s t the set and 10 microseconds l a t e r the reset*, Only 1 of the 16 output modules at the top of the c i r c u i t i s enabled by the. multiplexor and so only the selected l i n e w i l l be affected.. After issuing the desired command to an output l i n e an a d d i t i o n a l command should be issued to the same li n e (Dal2 and Dal3 both zero) causing both the set and reset l i n e s to be placed i n a low l o g i c state... The general input/output module serves as the l i n k between the Interdata minicomputer and the baseline compensators and the i n l e t system c o n t r o l l e r . I t s position in the system i s discussed more f u l l y in Chapter V.. 69 3.6 High Voltage Beam Suppression C i r c u i t r y l o accomplish baseline zeroing of the parametric amplifiers the ion beams must be deflected to prevent them from entering the Faraday cups._ Although i t was i n i t i a l l y intended to deflect the beams by applying a d i f f e r e n t i a l voltage to the Y plates of the Loveless source described i n Chapter I I . . Implementation of high voltage beam suppression entailed designing a simple d i g i t a l c i r c u i t that activated or deactivated one or another of two pairs of high voltage reed relays, which were connected to the Y plates. . In one configuration the voltage on the Y plates was controlled by potentiometers adjustable from the front panel. When the other relays were activated the voltage on the plates was determined by an i n t e r n a l potentiometer setting chosen so that no beams would enter the Faraday cups.. The c i r c u i t could be triggered either manually or under computer control. The: c i r c u i t r y was constructed and i n s t a l l a t i o n was routine. However, when the def l e c t i n g voltage: was removed the high voltage took more than 30 s to return to i t s steady state value.. This phenomenon was also noticed by Bi.Ryan who was using a Loveless type source in a s o l i d source lead mass spectrometer i n our laboratory. The Loveless source.is constructed such that the ions are accelerated through a potential difference of about 5 V 70 from where they are created to where they enter the Y deflectors.. as the ions traverse t h i s gap they are. moving at a r e l a t i v e l y low velocity of 4.5x103 km/s compared to v e l o c i t i e s about 30 times greater in the .beam tube. . The long time constant i s most l i k e l y due: to charging of an in s u l a t i n g surface i n the source when the beam i s deflected.. The "slow" moving ions are quite susceptible to the high voltages present on the charged surfaces and are deflected away from optimum source t r a j e c t o r i e s . as the charge bleeds away the beam returns to i t s optimum strength slowly. . I t would seem more.reasonable to design ion sources in a manner that accelerates them to high velocity as soon after they exit the i o n i z a t i o n chamber as possible._ I t should be noted that i f no variations are made to source potentials, the Loveless source behaves s a t i s f a c t o r i l y . . For the reasons discussed i n the preceding paragraphs, beam supression was accomplished by varying the magnetic f i e l d * . For the period of time that samples are being analyzed, no adjustments are made to the potentials present in the.Loveless source. 7 1 IV. ,THE DUAL SIDED INLET SYSTEM 4.1 General Description The most v i s i b l y evident contribution of t h i s thesis project i s the automated, dual sided i n l e t system.. This system i s comprised of three separate parts: the s t a i n l e s s s t e e l i n l e t system i t s e l f , the microprocessor based i n l e t system c o n t r o l l e r * and the Toepler pump c o n t r o l l e r . . The i n l e t system i s a state . of the art f a c i l i t y that makes introducing samples to the mass spectrometer extremely simple. The actual i n l e t system i s depicted in Figure 4.].. I t consists of 19 e l e c t r i c a l l y operated magnetic solenoid valves (Skinner Valves part number B2DX70).. The valves are interconnected using 0.25 inch annealed s t a i n l e s s s t e e l tubing and Swagelok s t a i n l e s s s t e e l f i t t i n g s * Connections to the valve bodies are made by coating Cajon 0.125 inch male pipe thread conversion f i t t i n g s with Araldite -. epoxy and screwing them into the o r i f i c e s on the Skinner valve bodies. The pressure transducers depicted on the: diagram are available from National Semiconductors as part number LX1702A. These transducers measure pressure, by amplifying the p i e z o e l e c t r i c voltage generated across a s i l i c o n c r y s t a l * . C i r c u i t r y , i n t e r n a l to the transducers, produces a very low impedance voltage output that varies l i n e a r l y with FIGURE 4 . 1 THE DUAL SIDED INLET SYSTEM The valve . configuration for the automated s t a i n l e s s s t e e l i n l e t system i s depicted i n the following figure.. Valves 1 and 11 are detachable and provide the means of mounting standard and unknown samples.. The use of a single gas admittance leak into the. mass spectrometer i s possible due to the extremely low volume of the portion of the i n l e t system between valves 9, 10 and 19.. TO MASS SPECTROMETER TOEPLER PUMP TOEPLER PUMP 74 pressure. The particular sensors used have a dynamic range of vacuum to atmospheric pressure with a s e n s i t i v i t y of 13mV/torr and a r e p e a t a b i l i t y of better than 3 t o r r . . These s o l i d state pressure transducers would be : very useful i n many mass spectrometer applications* The solenoid valves have been leak tested i n our laboratory and have been found to be leaktight below the 10-* t o r r l e v e l for a period of 6 hours. , Since the lowest pressure of in t e r e s t i n the i n l e t system i s about 20 t o r r , these valves are very acceptable. , At the time of t h i s writing the valves cost $25 each and t h i s low price adds to their d e s i r a b i l i t y . The Toepler pumps are large volume reservoirs which contain 500 cm3 of mercury.. The l e v e l of mercury within these pumps can be controlled and therefore the pressure of sample.gas above the reservoirs i s c o n t r o l l a b l e . . The 1000 cm3 reservoirs provide a large storage buffer to help diminish the e f f e c t of loss of sample gas through the day... The low volume area between valves 9 and 19 serves as a common gas admittance area to the stainless s t e e l c a p i l l a r y tubing that serves as the i n l e t leak to the mass spectrometer*. When sample sides are altered t h i s low volume area i s f i r s t evacuated to a suitably low pressure and then the valves are. cycled to admit gas from the other side of the i n l e t system to the mass spectrometer.. The entire assembly i s mounted on an aluminum back plate which serves as a heat sink for the valves as well as 75 providing a means of physical support.. 4 high vacuum system consisting of a rotary pump, an o i l d i f f u s i o n pump and a molecular sieve f o r e i i n e trap are mounted on the rear of the back plate. This i s the f i r s t use of an a l l metal i n l e t system i n our laboratory*„ I n i t i a l t e s t s have shown that the use of sta i n l e s s s t e e l has no disadvantages.. The system i s e a s i l y leak tested by exposing d i f f e r e n t areas of the i n l e t system to acetone. Leaks are very evident by monitoring pressure fluctuations with a thermocouple gauge. When a leak i s found i t i s ea s i l y fixed by tightening the appropriate Swagelok connection while the i n l e t system remains evacuated. It i s my experience that s t a i n l e s s s t e e l i n l e t systems are.a better choice than glass systems i n every aspect except cost.. When one:considers the;downtime that glass systems possess, the i n i t i a l l y higher cost of sta i n l e s s s t e e l becomes inconsequential. Due to the segmented construction of the.sample l i n e , components that f a i l i n time can e a s i l y be replaced.. It i s unlikely that downtime for component replacement would exceed one day. . Figure 4.2 shows a cross-section of a Toepler pump. The assembly consists of two separate parts. The large diameter cylinder at the bottom of the pump i s a mercury reservoir that i s large enough to contain a l l the mercury i n the pump.. a Cajon u l t r a - t o r r f i t t i n g connects the top of the mercury reservoir to a set of valves controlled by the FIGURE 4.2 CROSS-SECTIONAL VIEW OF THE TOEPLER PUMPS DESIGNED FOR THE AUTOMATED SYSTEM The following figure i s a scale drawing of the o n e - l i t e r -Toepler pumps designed for the automated mass spectrometer system., A l l f i t t i n g s are u l t r a - t o r r f i t t i n g s manufactured by Cajon. The volume of the Toepler pumps i s 250 cm3 from the centering electrode to the check assembly. The viewing tubes are.6 mm pyrex tubing. . Each Toepler pump contains 500 cm3 of mercury. TO MASS SPECTROMETER VIEWING TUBE L O W - -V A C U U M R C U R Y E N T R A N C E -H O L E S CHECK A5SEMBLY S A M P L E CARBON DIOXIDE H3- CENTERING ELECTRODE S T A I T - t E S S S T E E L C H E C K B A L L -ADJUSTABLE HEIGHT MERCURY COLUMN TO CONTROL \ * \ L V E S . SEE FIGURE * . 5 i t MERCURY RESERVOIR 78 Toepler pump c o n t r o l l e r . . These.valves can either evacuate a i r from, or bleed a i r i n t o , the top of the reservoir to control the height of the mercury column i n the sample cylinder.. The bottom of the reservoir i s shaped such that mercury i s directed toward the four holes in the bottom of the sample cylinders The mercury enters the :column; through these holes* The sampleicylinder i s paralle l e d by a glass tube which enables one to vi s u a l l y check the height of the mercury column.. An electrode i s situated halfway along the, volume and enables the Toepler pump co n t r o l l e r to center the mercury column before a sample: i s allowed to enter the Toepler pump. , The connection to the mass spectrometer i s at the top of the pump through a Cajon u l t r a - t o r r f i t t i n g . . . A check b a l l mating assembly i s situated at the top of the sample column. A st a i n l e s s s t e e l b a l l bearing, f l o a t i n g on the top of the mercury column, acts as a check b a l l in the event of a sudden change i n mercury l e v e l and prevents any mercury loss from the pump. The Toepler pumps are capable of producing volume changes of 500 cm3., Used in conjunction with the 1000 cm3 reservoirs, these pumps have no d i f f i c u l t y i n maintaining proper gas flows into the mass spectrometer for periods up to 4 8 hoUrs after they have been charged with carbon dioxide. 79 4.2 The Toepler Pump Controller The most time consuming step in the precise determination of stable isotope r a t i o s i s often adjusting the rate at which gas flows into the mass spectrometer.. Ahern (1975a) gives r e s u l t s of a multiple regression analysis of mass spectrometer parameters and concludes that the voltage c o e f f i c i e n t of resistance of the.. Victoreen r e s i s t o r s used i n the measuring system was the largest source.; of error in isotope r a t i o determinations (Whittles, (I960). This source of error can most easily be handled by making certain that the output voltage f o r the mass 44 ion beam always f a l l s within a predetermined l i m i t . , The Toepler pump c o n t r o l l e r i s a c i r c u i t constructed to regulate the magnitude of the output voltage for either the mass 46 or mass 44 ion beams. The operator d i a l s the desired peak height on a thumbwheel switch and also chooses a tolerance voltage.„ The Toepler pump con t r o l l e r ' s function i s to i n t e l l i g e n t l y open and close valves which control the l e v e l of the mercury in the Toepler pumps.. This i n turn alters the in t e n s i t y of the ion beam at the c o l l e c t o r * . Previous experience gained by analyzing samples helped to i d e n t i f y the design c r i t e r i a for the Toepler pump co n t r o l l e r . . T h e . c r i t e r i a were as follows: 1. . The Toepler pump c o n t r o l l e r had to be able to center the mercury i n the Toepler pumps.. Otherwise a series of samples that were either top large or too 80 small would cause the co n t r o l l e r ' s operating capacity to be exceeded.. 2. _ The c o n t r o l l e r had to be capable of either increasing or decreasing the sample gas flow into the mass spectrometer. 3., The time that valves are opened for a l t e r i n g mercury le v e l s should be functionally related to . the difference between the actual peak height and i t s desired value. . 4.,. After a l t e r i n g the mercury l e v e l the c o n t r o l l e r should delay a time proportional to the length of time determined i n the above . step. This delay allows the flow into the mass spectrometer to reach a steady state. 5. , The Toepler pump c o n t r o l l e r should continue adjusting the mercury l e v e l i n the: above fashion u n t i l the peak height of in t e r e s t f a l l s within a window defined by the settings of two thumbwheel switches on the front panel.:; I t was decided to design and build the Toepler pump co n t r o l l e r using integrated c i r c u i t s from the : Low Power Schottky TTL family. The f i n a l design produced a c o n t r o l l e r that s a t i s f i e d a l l of the above c r i t e r i a and reduced the time needed for peak height equalization from an average of about 10 minutes to les s than 30 seconds. The c o n t r o l l e r l o g i c a l l y consists of three separate portions: an . analog computation board, a d i g i t a l l o g i c 8 1 board, and a combination l e v e l centering and driving board. In addition to these c i r c u i t s , the:controller also possesses a multiplexed d i g i t a l voltmeter so that the i voltages of points of i n t e r e s t can be rea d i l y monitored.. Each of these c i r c u i t s w i l l be discussed b r i e f l y . . 4.2.1 The Toepler Pump Analog Computer Since the ion beams e x i s t i n the analog domain and the c o n t r o l l e r functions d i g i t a l l y , the analog board was necessary to provide an interface :between the two domains. Figure 4.3 shows the analog c i r c u i t r y used i n t h i s Toepler pump c o n t r o l l e r . The center voltage (A) i s defined by a 4 d i g i t Kelvin-Varley divider. The output of the: divider i s buffered by an OP-07 u l t r a low o f f s e t operational amplifier to insure the l i n e a r i t y of the : divider,. The output of a 3 d i g i t Kelvin-Varley divider i s buffered by another OP-07 amplifier to define the window voltage (B) . Both of these dividers are adjustable by thumbwheel switches located on the front panel of the c o n t r o l l e r . . These two outputs are then combined in an inverting summing amplifier •to produce - (A+B) i n the middle of t h e . c i r c u i t , and combined i n an inverting difference amplifier to produce - (A-B).. The outputs of the summing and difference amplifiers are connected to the inverting inputs of the. two precision FIGURE 4.3 THE ANALOG COMPUTER CIRCUIT OF THE TOEPLER PUMP CONTROLLER The analog computer c i r c u i t computes the maximum and minimum peak voltages that are acceptable for a mass spectrometer peak height to be a correct magnitude. The information i s dist r i b u t e d to the: rest of the Toepler pump c o n t r o l l e r by the outputs of the two comparators i n the upper right of the c i r c u i t . . The absolute value of the difference between the actual peak height and the desired peak height i s also calculated by this c i r c u i t . _ CI C2 . WINDOW VOLTAGE 3 DIGIT IKELVIN VARLEY DIVIDER . CENTER VOLTAGE 10 V 4- DIGIT KELVIN VARLEY DIVIDER • ERROR VOLTAGE PEAK VOL TAW BUTTERWORTH FILTER BUFFER 0PO7^ BUFFER OP07> r-A 5-lA-Cl - vVA -DIFFERENCE OPOT: -15 -vW 1 SUM 0P07^ WV R |A-C| 1 CI R -IA-BU (A4-BI R •VAr r-AAAq -< i < > • i H •AAA-COMPARATOR COMPARATOR kMPO? r-W-MEASURING SYSTEM GROUND A B S O L U T E V A L U E CIRCUITRY R , •vWM 1 COMMON MODE REJECTION AND INVERTER 84 comparators. The peak height of i n t e r e s t enters the board near the bottom of the c i r c u i t and i s Butterworth f i l t e r e d to remove a large 60 Hz noise component.. The output of the f i l t e r i s connected to the inverting input of a difference amplifier*. The non-inverting input of t h i s difference.amplifier i s connected to the measuring system ground.. This method of introducing the peak voltage (C) into the Toepler pump co n t r o l l e r was necessary to remove common mode:noise i n the measuring system as seen by the c o n t r o l l e r . : The inverted peak voltage i s Connected to the non-inverting inputs of the previously mentioned comparators.. The outputs of the comparators (CI and C2) d i g i t a l l y encode the r e l a t i o n of the peak height to the center and window voltages mentioned previously.. Table 4.1 indicates the comparator's outputs under various conditions. . 85 CONDITION COMPARATOR U COMPARATOR 2 i C<A-B 1 low 1 low OA+B ! high 1 high A-B<C<A+B ! high 1 low impossible low high TABLE 4.1 THE POSSIBLE COMBINATIONS OF PEAK HEIGHT COMPARISONS The main function of the Toepler pump c o n t r o l l e r i s to i n t e l l i g e n t l y a l t e r the mercury l e v e l in the:Toepler pumps. Instead of just opening valves appropriately to increase or decrease the peak height, the c o n t r o l l e r i s also i n t e l l i g e n t enough to open valves f o r longer periods of time i f the peak height i s f a r from i t s desired s e t t i n g . . S i m i l a r l y i f the peak i s almost the correct height, the c o n t r o l l e r w i l l only open the valves b r i e f l y . . To accomplish t h i s task i t i s necessary for the c o n t r o l l e r to calculate the absolute.value of the difference between the peak height, C, and the desired voltage, A. The two amplifiers in the lower right of Figure 4.3 are configured i n such a manner that t h e i r output w i l l be equal to I Peak Height-Center Voltage|.. [ 4 - 1 ] The d e t a i l s of the absolute value c i r c u i t w i l l not be 86 discussed here but i t w i l l be said that the diodes obviously play the c r u c i a l role. . The f i n a l 74V operational amplifier produces an output voltage equal to 5-| C- A | volts [4.2] which i s passed onto the d i g i t a l l o g i c board discussed i n the nsxt section. 4.2.2 The Toepler Pump D i g i t a l Logic Board The Toepler pump d i g i t a l l o g i c board provides the hardware i n t e l l i g e n c e . that determines when the valves controlled by the Toepler pump c o n t r o l l e r are activated or deactivated. The logic c i r c u i t for t h i s board i s presented in Fig ure 4.4. A s t a r t s i gnal, from either a button press or the computer input, enters the c i r c u i t by the NAND gate in the far r i g h t of the c i r c u i t which i n turn clocks a D type f l i p - f l o p . As the high l o g i c l e v e l appears two things happen concurrently: the toggle out control pulse i s produced and sent off the board to the driver board causing the matched control l i n e to be set properly.. At the same time, the leftmost multivibrator i s triggered enabling several gates.. At t h i s time the 3 EX AB 2240 programmable counters are reset. I f the peak i s already matched at this point, the input FIGURE 4.4 THE TPC DIGITAL LOGIC CONTROL BOARD This c i r c u i t provides the fundamental control l o g i c necessary to drive the Toepler pumps i n t e l l i g e n t l y * . The upper l e f t portion of the c i r c u i t i s a d i g i t a l memory used to store an 8 b i t number proportional to the error voltage. The ac t i v a t i o n and delay times are.proportional to t h i s 8 b i t number. E R R O R V O L T A G E C O M P A R A T O R > R C T I M E R nc T I M E R DAC XR2240 A C T I V E T I M E ICOUNTERl X R 2 2 4 0 D E L A Y T I M E COUNTER| R DIGITAL MEMORY \ — \ B OUTPUTS XR2240 DIGITAL M E M O R Y R C TIMER a B IT DIGITAL ICOMflARATOR O U T P U T HIOH W H E N I N P U T S E Q U A L RESET OICITAL ERROR VOLTAGE . TOGGLE OUT . E N A B L E L E D „ <2 MANUAL START D I S A B L E L E O ^ 1 T W A T C H ^ a B I T DIGITAL COMPARATOR O U T P U T HIGH W H E N I N P U T S E O U A L 1 M A T C H L E D ^ JL ' - Y 1 U N M A T C H E D L E D ^ 00 00 89 D t y p e : f l i p - f l o p w i l l be reset and the device recognizes that i t s task i s completed.. If the peak i s not matched s a t i s f a c t o r i l y , the second monostable vibrator w i l l be triggered which resets latches and triggers the uppermost programmable counter. The output bus of this counter i s attached to a d i g i t a l to analog converter whose output i s compared with the error voltage computed on the analog board discussed i n the preceding section.. The-output of the operational amplifier connected to the D/A converter increases in a stepwise fashion u n t i l i t s output exceeds the. error voltage.. When th i s occurs the output of the .precision comparator changes from a low to a high state, clocking the bistable l a t c h . The complementary output of the .latch goes low, disabling the programmable timer, while the Q output goes high trig g e r i n g the ACTIVE programmable counter and generating the 22 40 OK signal connected to the driver board.. The outputs of the uppermost programmable counter remain s t a t i c and the device functions as an 8 b i t d i g i t a l memory. The ACTIVE programmable counter counts at a rate that i s determined by a c i r c u i t capacitor and a potentiometer that i s adjustable from the front panel. The outputs of the ACTIVE counter are l o g i c a l l y compared with the 8 b i t number stored in the d i g i t a l memory. When they are equal the output of the 8 b i t d i g i t a l comparator clocks a D type latch disabling the ACTIVE counter and resetting the 2240 OK signal passed off the board. , The latch also triggers the 90 WAIT programmable timer.. The WAIT programmable counter functions i n a s i m i l a r manner to the ACTIVE counter. Its count rate i s controlled by a second potentiometer on the front panel. When the 8 bi t value of the WAIT counter i s l o g i c a l l y equal to the error voltage stored i n the d i g i t a l memory* the. wait cycle i s at an end. Depending on whether or not the peak i s now matched one of two things w i l l happen. I f the peak i s matched to the settings of the thumbwheel switches then the start D type la t c h i s reset and the c i r c u i t ' s task i s completed.. However, i f the peaks are unmatched, a new cycle w i l l begin much as i f a second start pulse: had been detected,. As i s now evident, the:length of time that a valve i s opened i s determined by the magnitude of the error voltage (the size of the:8 b i t number in the d i g i t a l memory) and the count rate of the ACTIVE counter.. In fact, the length of time the valves are activated i s d i r e c t l y proportional to the error voltage and can range from 1 to 256 clock cycles. Error voltages between 0 and 5 V can be accommodated.. The delay c i r c u i t r y functions in a similar fashion. The delay period i s also proportional to the error voltage. . This allows the peak height i n question to reach a steady state after being altered during the activation period* By adjusting the potentiometers on the front panel a matched condition can generally be obtained with a single i t e r a t i o n of the Toepler pump c o n t r o l l e r . . 9 1 4.2.3a The Toepler Pump Driver Board The c i r c u i t r y described i n the two previous sections determines when valves should be opened and when they should be closed.. The Toepler pump driver board provides the:logic that determines which valve (s) should be opened. . Figure 4.5 shows the valve configuration attached to the l e f t and right Toepler pumps. I f the l e f t Toepler pump i s to be adjusted then valve 2 i s activated.. Valve 3 i s activated for adjustments to the right Toepler pump. If valve 1 i s not energized then atmospheric pressure w i l l flow into the selected pump. Activating valve 1 w i l l connect a vacuum pump to the selected Toepler pump and the mercury l e v e l in the pump w i l l decrease appropriately. The valve configuration makes i t impossible to connect the vacuum to high pressure. .  The l o g i c on the driver board i s very straightforward and s t i l l provides f a i l safe operation. . The toggle out control l i n e , generated when the st a r t s i g n a l i s detected, clocks the outputs of the two analog comparators into the two D type latches i n the lower portion of Figure 4.6. , These latches are not clocked again u n t i l a new cycle i s i n i t i a t e d and so they serve as a memory as to the state of the peak height at the beginning of the cycle., Contrary to t h i s , the matched signal generated by the two EXCLUSIVE NOR gates i s a dynamic indicati o n of whether the peak i s matched or not* The two FIGURE 4.5 CONTROL VALVE CONFIGURATION FOR THE TOEPLER PUMPS This diagram shows the positions which the 3 valves controlled by the Toepler pump c o n t r o l l e r occupy r e l a t i v e to the actual Toepler pumps. The adjustable viscous leaks provide a means to l i m i t the gas flow through any valve(s) and therefore are important i n determining the period of time valves must be activated to af f e c t the peak heights i n a satisfactory manner. TO MASS SPECTROMETER TO MASS SPECTROMETER STANDARD S I D E T O E P L E R P U M P A I R PRESSURE ADJUSTABLE VISCOUS LEAK 3-WAY VALVE, N.O ADJU STABLE N.C. 2-WAV VALVE VISCOUS LEAK 730 TORR VACUUM 2-WAY VALVE N-C. N.C. U N K N O W N SI DE T O E P L E R PUMP U> FIGOBE 4.6 THE TOEPLER POMP DRIVER BOARD The l o g i c necessary to open and close the valves shown i n Figure 4*5 i s shown i n t h i s f i g u r e . . In addition, the c i r c u i t r y shown i n the upper portion of the figure i s responsible for the recentering of the mercury within the Toepler pumps whenever a centering command i s received. 96 sets of four EXCLUSIVE NOR gates in the bottom center of Figure 4.6 act as 4 b i t comparators. For the driver board to request a lowering of the mercury l e v e l , four conditions must be s a t i s f i e d ; both comparators must be in a high state, the busy b i t from the logic board must be high and the ACTIVE 2240 si g n a l from the c i r c u i t discussed in the previous section must be i n a low state. , Similar conditons must e x i s t for an increase i n mercury l e v e l with the exception that the two comparators must be i n a low state. The rest of the driver board c i r c u i t r y i s e a s i l y understood.. I t provides l o g i c which combines requests for increases and decreases i n mercury l e v e l changes which may come from the c i r c u i t r y discussed above,, from manual requests generated by pressing buttons on the front panel, and from requests generated by the mercury centering c i r c u i t r y in the top center of Figure 4.6. A busy b i t i s generated by the 5 input OR gate i n the upper portion of the figure. . Performance of any task w i l l cause the busy b i t to go high. A l l electromagnetic valves and most l i g h t emitting diodes are driven by 9667 PMOS drivers situated on the driver board. These drivers e a s i l y drive the 0.3 A required by each valve and come i n a convenient dual i n l i n e package (DIP) containing 7 drivers. . 97 4.2.3b The Mercury Level Centering C i r c u i t r y I f the quantity of carbon dioxide introduced into the Toepler pumps was always either i n s u f f i c i e n t or i n excess, the Toepler pump c o n t r o l l e r would cause the;mercury l e v e l to always change in the same d i r e c t i o n . After some time the mercury would be at one of i t s extreme positions and the co n t r o l l e r would be unable to complete a request for peak equalization. However, i f the mercury l e v e l i s returned to a position near the center of i t s operating range before carbon dioxide i s allowed to enter i t , t h i s problem i s all e v i a t e d . The c i r c u i t which performs centering of the mercury column i s i n the upper portion of Figure 4.6. . The mercury in the;Toepler pump i s used as a mercury switch to determine whether the mercury should be lowered or raised* The four D type latches function i n two pairs, two latches for the right hand Toepler pump and two for the l e f t * , The mercury switch generates either a high or low d i g i t a l signal.„ One of the two latches w i l l be held reset while the other one w i l l be clocked and generate a request for either r a i s i n g or lowering the mercury column.e When the . mercury switch changes state as the mercury flows past the centering electrode, the la t c h generating the request w i l l be reset and the centering action w i l l be completed.... The centering feature of the Toepler pumps i s i n i t i a t e d 98 by separate control l i n e s to the :controller when the request i s generated by the computer* In the case of a manual request to center, the centering takes place on the side indicated by the. position of the side s e l e c t switch. Requests for peak equalization are always made on the side flowing into the mass spectrometer when the side s e l e c t switch i s i n the normal AOTO position. 4.2.4 The Design of a Multiplexed D i g i t a l Voltmeter The complex c i r c u i t r y contained within the Toepler pump con t r o l l e r made i t advantageous to monitor the analog voltages at several key points i n the c i r c u i t . The recent a v a i l a b i l i t y of single chip d i g i t a l voltmeters made i t reasonable to place the analog to d i g i t a l converter within the Toepler pump co n t r o l l e r . In the Toepler pump c o n t r o l l e r i t i s necessary to be able to measure the r e l a t i v e voltage differences of six points within the c i r c u i t . These voltages are: the peak height^ the error voltage computed on the analog computer board, the d i g i t a l sample and hold error voltage, the center voltage, the window voltage, and analog ground. The a b i l i t y to measure r e l a t i v e differences between such points as peak voltage and center voltage and to compare t h i s difference with the error voltage allows the operator to guickly v e r i f y proper functioning of many d i f f e r e n t portions of the 99 c i r c u i t . The existence of d i g i t a l l y controlled analog multiplexors prompted their i n c l u s i o n i n the design of a multiplexed d i g i t a l voltmeter for the Toepler pump Controller. The 8 analog input l i n e s l e f t room for the i n c l u s i o n of two banana plug connectors on the front panel. This acts as a b u i l t - i n d i g i t a l voltmeter in the mass spec tro meter* Figure 4.7 i s a l o g i c diagram which shows the c i r c u i t r y b u i l t for the multiplexed d i g i t a l voltmeter. The 8 push button switches shown i n the upper l e f t of the: c i r c u i t are debounced by 8 E/S f l i p - f l o p s and input into an 8 input encoder.. Depending on whether the positive or negative input has been selected by a two position toggle switch, the three output l i n e s of the encoder are latched into one.of two latches. A 3 b i t comparator samples the inputs and outputs to the latch and when equal, places a high on the active low enable input of the latches, disabling the latches* The l a t c h outputs are attached to the channel selector inputs of an 8 channel analog multiplexor, causing the selected input voltage to be connected to the output of the appropriate multiplexor. The output of the lower multiplexor i s connected d i r e c t l y to the negative input of the d i g i t a l voltmeter as well as one input of a r e s i s t o r attenuator. ,. The output of the upper multiplexor i s connected to the r e s i s t o r attenuator where i t i s reduced by some power of ten (set by 100 FIGURE 4.7 THE TOEPLER PUMP MULTIPLEXED DIGITAL VOLTMETER CIRCUIT The b u i l t i n d i g i t a l voltmeter i n the Toepler pump c o n t r o l l e r is' multiplexed i n a manner allowing any possible combination of 8 input voltage signals to be attached to the positive and negative inputs of the voltmeter* . Three decades of attenuation make i t possible to measure signals as large as +75 V and as small as -75 V. ,. PEAK DSH DYN CEN WIN A A A A A A A A SWITCH DEBOUNCING CIRCUITRY 2X4.0*3 + riNPUT SELECT 8:3 ENCODER 4-532 r - 3 BIT -COWWE -- 40083 - 3 BfT - COWWlEf-- 400BS I 3 BIT 1 LATCH f 40*2 I E L A 8 CHANNEL ANALOG MULTI-PLEXOR 4051 B I CHANNEL ] | ANALOG MULTI-PLEXOR 4051 A T T E N U A T O R INTERSIL 710* RANGE SELECT 102 the range control switch) before entering the positive input of the. d i g i t a l voltmeter. The multiplexed voltmeter can measure; any voltage between -75 V and +75 V with 3 1/2 d i g i t resolution.. The inputs are f l o a t i n g and possess an input impedance of 1x106 ohms. The DVM i s available from I n t e r s i l as part number ICL-7106.. The device has proven to be extremely v e r s a t i l e but i t s major use i s as a monitor for the mass 44 and mass 46 peak heights.. I t has also proven very useful i n the debugging and i n i t i a l set up of the entire system. 4.3 The Microprocessor Based Inlet System Controller 4.3.1 General Description In the early stages of the system design we recognized that a sophisticated i n l e t system, and associated co n t r o l l e r , could greatly reduce the time: required to introduce hew samples of carbon dioxide into the mass spectrometer. The s t a i n l e s s s t e e l i n l e t system discussed i n section 4,1 was designed with rapid sample handling as a p r i n c i p a l c r i t e r i o n . When the c o n t r o l l e r for the i n l e t system was i n i t i a l l y being designed i t was clear that a simple 103 sequencing c o n t r o l l e r could c e r t a i n l y perform the task but such a system would never function as i n t e l l i g e n t l y as a technician.. Preliminary discussions with. R.D. Meldrum concluded with a decision to build the: i n l e t system c o n t r o l l e r using a microprocessor. The i n t e l l i g e n c e that could be designed i n t o such a system was considerable, and could e a s i l y out perform a technician for t h i s s p e c i f i c task.. The design of the hardware portion of the microprocessor based i n l e t system c o n t r o l l e r was undertaken by R.D.,Meldrum., The c i r c u i t was customized to the s p e c i f i c requirements of the i n l e t system. I t i s my experience that this customization i s where the f u l l power of a microprocessor c i r c u i t l i e s . , Applications of a microprocessor in the form of an evaluation board normally make the microprocessor function as a microcomputer and cannot take f u l l advantage of the microprocessor.. The extended family of peripheral integrated c i r c u i t s for most microprocessors greatly simplify the hardware design. The microprocessor chosen for t h i s application was the Motorola 6802 which comes complete with a 4 MHz clock and 128 bytes of random access memory on the same chip. . This choice. was made primarily because of the. experience R.D. Meldrum had obtained in a previous application. This choice.seems to have been well made since the Motorola family of microprocessors i s progressing at a much fas t e r rate than processors available from any other manufacturer.. 104 The recent a v a i l a b i l i t y of the Motorola 6809 and Motorola 68000 microprocessors emphasises t h i s point extremely well (Hartman, 1979; Gooze, 1979).. Figure 4.8 i s a block diagram showing the hardware configuration of the microprocessor based i n l e t system controller... The co n t r o l l e r consists of a Motorola 6802 microprocessor, 128 bytes of random access memory (RAM), four kilobytes of read only memory, i n the form of two In t e l 2716 programmable read only memories (PROMs), and four Motorola 68 21 peripheral interface adapters (PIAs) which l i n k the system with the external world._ The only other hardware in the i n l e t system c o n t r o l l e r i s related to chip selection or solenoid or l i g h t emitting diode:driving. , The e n t i r e . c i r c u i t r y for the c o n t r o l l e r consists of fewer than two dozen integrated c i r c u i t s . The complexity of the co n t r o l l e r i s i n the software* not the : hardware. This feature i s one of the more important advantages of microprocessor based c i r c u i t s . The four sets of captions at the top of the figure indicate the various external devices to which the system i s interfaced.. These;include 19 valve buttons, 19 valves, 5 control buttons, 2 operator response buttons, 27 l i g h t emitting diodes, 2 LX1702 pressure transducer c i r c u i t s , 4 Toepler pump c o n t r o l l e r control l i n e s , 4 sample preparation l i n e - control l i n e s , 4 minicomputer control l i n e s and 1 audio tone .generator.. This p r o l i f e r a t i o n of input/output l i n e s gives the f i r s t i n d i c a t i o n of the complexity of the control l e r * FIGURE 4.8 THE BLOCK DIAGRAM OF THE MICROPROCESSOR BASED INLET SYSTEM CONTROLLER The s i m p l i c i t y of the hardware surrounding the microprocessor used i n the i n l e t system c o n t r o l l e r i s c l e a r l y shown i n t h i s block diagram. . The 4 PIAs (Peripheral Interface Adapters) provide the in t e r f a c i n g between the microprocessor and the outside world., Only 1/2 of the 4K bytes available i n the PROMs are used by the control program but t h i s allows easy modification and expansion of the i n l e t system c o n t r o l l e r i f deemed necessary i n the future* CONTROL PIA H\ I. AUDIO TONE GENERATOR I CONTROL BUTTONS a OOO BUTTONS 1.10 II 4. OOO V A L V C 6 1*10 + II 5. INTERDATA MINICOMPUTER CONTROL PIA #2 ' I. TOEPLER PUMP CONTROLLER i SAMPLE LINE CONTROLLER 5 CONTROL BUTTONS 4, 10 HZ TIMER SAMPLE SIDE PIA t. SAMPLE BUTTONS 2. SAK*1_E VALVES 1 STANDARD PRESSURE TRAMS. STANDARD SIDH PIA I. STANDARD BUTTONS 1. STANDARD VALVES 1 SAMPLE PRESSURE TRAMS 4K EPROM I2B B RAM ON F3OAR0 MCC802 * MHZ CRYSTAL 01 CLOCK lb B ' T A P P R E 5 S BUS MICRO PROCESSOR RF A D / W R I T E -VALID MEMORY A0DRF5S INTERRUPT REQUEST i 8 BIT BIDIRECTIONAL DATA 0U5 N—n 107 The p r i n c i p a l task of the i n l e t system c o n t r o l l e r i s to open and close valves, in proper sequence, to introduce new samples to the mass spectrometer. Functions such as evacuation of selected portions of the i n l e t system and the movement of samples from one:area i n the i n l e t system to another, are under the control of the microprocessor. The basic sequence of steps that must be performed to introduce a sample into the:mass spectrometer can be most easily understood by r e f e r r i n g to Figure 4.1._ An aliquot of tank carbon dioxide i s attached to the i n l e t system by mounting valve 1 using a high vacuum f i t t i n g . . The volume beween valves \, 2 and 5 i s evacuated.. Valves 2 and 3 are opened and carbon dioxide flows through the: leak into the 25 ml reservoir.. When the pressure i n the LX1702A pressure transducer exceeds some preset value, the: microprocessor closes valves to i s o l a t e the standard carbon dioxide i n the reservoir. . A command must be given to the Toepler pump co n t r o l l e r to center the: mercury i n the l e f t hand pump. When t h i s action i s completed, valves 4, 7 and 8 are opened to allow the. tank carbon dioxide to flow into the Toepler pump.. After a predetermined delay, valve 9 can be opened to admit the gas to the mass spectrometer* After some delay, a command can be given to the Toepler pump c o n t r o l l e r to equalize the l e f t hand side peak height* J This tank carbon dioxide serves as a working standard throughout the analysis day. The f i r s t unknown sample i s loaded in e s s e n t i a l l y the 108 same manner as the working standard. . To minimize delays between analysis of subsequent unknowns, i t i s desirable to have the next unknown sample present in the .2 5 ml reservoir before the analysis of the previous unknown i s c ompleted. Therefore after an unknown has been transf erred into the Toepler pump region, the 25 ml reservoir a rea i s fir,st evacuated of the previous sample and the next sample i s allowed to flow through the;leak into the rese r v o i r region. Upon completion of t h i s task a thi r d unknown can b e mounted and the volume between valves 11, 12 an d 15 can be evacuated. The i n l e t system c o n t r o l l e r performs the above tasks i n an i n t e l l i g e n t manner.. The control software i s written i n Motorola 6800 assembler language and assembled on the Amdahl 470/76-11 computer of the University of B r i t i s h Columbia's Computing Centre*, The assembly language program occupies 39 pages of source statements (see Appendix II) which generates over 2000 bytes of machine language code;. The program i s largel y interrupt driven and portions are reentrant. _,, The sophistication contained within the coding exceeds that of any program previously implemented i n the mass spectrometer lab. Most of the interrrupt service routines are themselves interruptable. The microprocessor has been programmed i n a " f a i l safe" manner. With only a few desirable exceptions, i t i s impossible to make the i n l e t system c o n t r o l l e r perform an operation that has previously been programmed as being undesirable* Operation in the manual mode removes most f a i l 109 safe features of the c o n t r o l l e r and permits a knowledgable operator to perform any task, even i f i t i s extremely uncoma on.. The c o n t r o l l e r functions by executing one of f i v e possible software routines. These five routines are; STANDBY, STANDARD LOAD, UNIVERSAL PUMP, MANUAL, and GO. Entry to the above modes of operation i s generally made by pressing the appropriate control button on the front panel with the exception of the GO routine, which may also be entered d i r e c t l y from STANDARD LOAD under program control.. 4.3.2 Standby, Manual and Universal Pump Control Modes »hen one i n i t i a l l y powers up the Inlet System Controller c i r c u i t r y , the c o n t r o l l e r i s placed i n STANDBY mode.. In t h i s configuration only MANUAL and UNIVERSAL PUMP are v a l i d selections and any other selection r e s u l t s in the microprocessor activating the audio tone generator which sounds as long as the button remains pressed. Pressing MANUAL allows any valve to be opened or closed by pressing the appropriate valve button on the front panel. Normally one presses the UNIVERSAL PUMP control button and the microprocessor responds by opening and closing valves in such a manner that a l l areas of the i n l e t system are connected to the waste vacuum.. After a delay of 10 110 minutes the microprocessor begins flashing the standard load l i g h t emitting diode, indicating to the operator that the universal pump time has expired and i t i s appropriate to enter the standard load routine. , 4.3.3 Standard Load Control Mode The STANDARD LOAD routine i s responsible for requesting that a sample be mounted on t h e : l e f t hand side of the i n l e t system. It i s also responsible for moving the standard from the re s e r v o i r below valve 1, into the Toepler pump and subsequently into the mass spectrometer i t s e l f . In the event that the amount of standard carbon dioxide i s i n s u f f i c i e n t , the standard load routine. w i l l sense the condition and return to the universal pump routine. After successfully t r a n s f e r r i n g the aliquot of carbon dioxide into the: mass spectrometer, the standard load requests that the operator optimize the potentials within the mass spectrometer to a point where s u f f i c i e n t s t a b i l i t y and magnitude of the ion beam are obtained.. The c o n t r o l l e r waits, p e r i o d i c a l l y a c t i v a t i n g the audio tone generator, u n t i l the operator presses the standard ready button. When the standard load routine senses that the mass spectrometer i s optimized i t responds by requesting that an unknown sample be mounted on the right hand side of the i n l e t system.. The sample request l i g h t emitting diode 111 remains l i t and the audio tone generator beeps p e r i o d i c a l l y u n t i l the sample ready button i s pressed. After evacuating necessary portions of the i n l e t system, the c o n t r o l l e r transfers the unknown carbon dioxide through a leak into the 25 ml reservoir. I f t h i s process i s not interrupted by the pressure transducer c i r c u i t r y before one minute has elapsed, the c o n t r o l l e r enters a low pressure f a i l routine.which takes the following action;. F i r s t i t pumps out a l l portions of the i n l e t system containing the sample causing the low pressure f a i l / second i t indicates the condition to the minicomputer over one of the interconnecting control l i n e s , t h i r d i t requests that a new sample be mounted, and fourth i t s t a r t s a c t i v a t i n g the audiotone generator at a frequency of 2 Hz, which i s an obvious indica t i o n to the.operator that a low pressure condition was encountered.. Normally the transfer of an unknown into the 25 ml reservoir i s interrupted by the pressure transducer c i r c u i t r y when an appropriate amount of carbon dioxide i s i n the reservoir*. The c o n t r o l l e r responds to the pressure transducer interrupt by closing the appropriate valves to i s o l a t e the carbon dioxide i n the 25 ml reservoir and by issuing a request for a new sample.. At t h i s point the unknown Toepler pump i s empty and so the i n l e t system c o n t r o l l e r begins transferring the carbon dioxide; from the 25 ml reservoir to the Toepler pump. As the carbon dioxide flows from the 25 ml reservoir area into the. Toepler pump, the i n l e t system c o n t r o l l e r i n i t i a t e s a procedure that 112 replaces the carbon dioxide flowing into the mass spectrometer from the standard side to the unknown side of the mass spectrometer i n l e t system (see next section). After a 5 s delay, the i n l e t system c o n t r o l l e r f i n i s h e s changing sides and closes valve 17, i s o l a t i n g the Toepler pump from the 25 ml reservoir.. The co n t r o l l e r then connects the 25 ml reservoir to the waste vacuum for 20 s and then issues a command to the Toepler pump co n t r o l l e r to equalize the peak height... The c o n t r o l l e r then enters the GO routine which i s the portion of the program where the majority of time i s spent* 4.3.4 The GO Control Mode The GO routine can be entered d i r e c t l y from the STANDARD LOAD routine or from the MANUAL, routine a f t e r loading the f i r s t unknown into the Toepler pump and pumping out the 25 ml reservoir manually.. The GO routine i s extremely involved due to the fac t that i t i s primarily interrupt driven causing the sequence i n which in s t r u c t i o n s are executed to vary considerably._ The GO routine services interrupts from the following sources l i s t e d i n order of p r i o r i t y : 1. „ the unknown side pressure transducer 2. „ a 10 Hz timer 3. _ the change sides control l i n e from the minicomputer 4. . the new sample request control l i n e from the 113 minicomputex 5. „ the sample ready switch 6 . _. the remaining push button switches The control functions performed by the GO routine are extremely d i f f i c u l t to describe i n d e t a i l due to t h e i f a c t that the sequence of i n s t r u c t i o n s executed i s determined by the order i n which interrupts are . processed by the microprocessor. , In addition the outcome of many conditional branches i s determined by the present "state" of the i n l e t system which varies considerably with time. . The state of the i n l e t system i s stored as a b i t encoded byte i n random access memory that can be thought of as f l a g s . A common programming technigue used in t h i s c o n t r o l l e r i s for the microprocessor to be executing a tight programming loop examining the state of one or more f l a g s . . The:processing of an interrupt a l t e r s the f l a g i n question and when the processor returns from the i n t e r r r u p t i t e x i t s the t i g h t loop due to a change i n the flag being examined.. To describe the operation of the GO control mode of the i n l e t system c o n t r o l l e r i n any d e t a i l would be : extremely d i f f i c u l t due to the role that interrupts play i n i t s functioning. I t i s f a r more reasonable to present the design of the programming in terms of two flowcharts depicting the l o g i c implemented i n the i n l e t system c o n t r o l l e r . . Figure 4.9a i s the flowchart showing the l o g i c i n the GO portion of the control program* One i s given the impression that the sequence of instructions i s e a s i l y FIGUBE 4.9a A FLOWCAHRT OF THE FUNCTIONING OF THE GO CONTROL MODE The normal operation of the i n l e t system c o n t r o l l e r i s i n the GO control mode.. The flowchart shown i n the:following figure provides a f a i r l y detailed description of the logic followed by the co n t r o l l e r while i n the GO mode.. The actual assembly language program that implements the l o g i c i n the flowchart i s provided in Appendix II;.. Flowchart symbols are those normally used by IBM with the exception that the large: ovals represent symbolic addresses used i n the assembly language program and act as transfer points that are usually represented by c i r c l e s . 1 1 5 116 traced in the flowchart, but i t must be remembered that an interrupt may occur at any point within the flowchart., When an interrupt i s detected by the microprocessor, control i s transferred to the point labeled ISE on Figure 4.9b* the flowchart showing the interrupt service routine.. Interrupt processing i s handled as shown i n Figure 4.9b. Frequently the sequence of instructions terminates with an RTI (Return from Interrupt) i n s t r u c t i o n i n which case control i s transferred back to the portion of the. program the processor was executing when the interrupt was received.. In some cases t h i s may be back to another portion of the interrupt service routine i t s e l f . . This i s permitted since most portions of the interrupt service routine are reentrant. . In at l e a s t two instances, an entry into the interrupt service routine does not terminate i n an RTI i n s t r u c t i o n but instead control i s transferred to the NEWSAM or MOVSAM entry points in Figure 4.9a.„ In these cases the stack pointer i s set to the top of the stack to prevent an eventual stack overflow condition and processing continues accordingly. The program flowcharts shown in Figures 4*9a and 4.9b provide nearly foolproof operation of the: oxygen mass spectrometer. As long as the c o n t r o l l e r remains i n the GO control mode and sample flasks.are r e a l l y mounted when the operator indicates one i s , by pressing the sample ready button, i t i s believed that i t i s impossible f o r the c o n t r o l l e r to make an error.. Operator intervention must meet c e r t a i n programmed c r i t e r i a before the c o n t r o l l e r w i l l FIGURE 4.9b THE FLOWCHART SHOWING THE INTERRUPT SERVICE ROUTINE DEVISED FOR THE INLET SYSTEM CONTROLLER The following flowchart shows the l o g i c used to process interrupts received while in the GO contr o l mode as well as a s i g n i f i c a n t portion of the STANDARD LOAD control mode. The RTI labels within the large ovals indicate that control i s to be returned to the point i n the program that was being processed when the interrupt was received.. 118 119 perform any given action and t h i s eliminates almost a l l sources of even operator error.. 4.3.5 The Inlet System Pump Routines One of the most sophisticated portions of the i n l e t system c o n t r o l l e r l i e s i n the manner in which the c o n t r o l l e r u t i l i z e s the waste vacuum system.„ At any given time more than one area of the i n l e t system may need use of the vacuum system., For t h i s reason, a method of pump u t i l i z a t i o n was developed that properly handled multiple requests for use of the waste vacuum system. Maximum sample throughput can be obtained by giving pump requests that are closest to the mass spectrometer leak the highest p r i o r i t y for the use of the pump.. For t h i s reason the pumping p r i o r i t i e s have been assigned in the following order: 1) the ; common area between valves 9, 10 and 19; 2) the sample Toepler pump and reservoir; 3) the sample 25 ml reservoir and i n l e t area; 4) the sample 25 ml reservoir; and 5) the sample i n l e t area between valves 11, 12 and 15. Requests f o r pumping i n the GO control mode are made through c a l l s to appropriate subroutines which insure that the waste vacuum i s connected to the portion of the i n l e t system with the highest p r i o r i t y . . I f the waste vacuum i s being used at the time of a second pump request, and the second request i s of a higher p r i o r i t y , the;pump routines 120 w i l l terminate the current pump, f l a g i t as an area that needs to be pumped as soon as the waste vacuum i s ava i l a b l e , and begin pumping on the second pump request. If the second pump request i s of a lower p r i o r i t y than the.current pump i t i s just flagged as pending* , Requests f o r pumping set the appropriate entry i n a delay table to a predetermined value.. Interrupts from the 10 Hz clock result i n each non-zero entry i n the table being decremented by 1.. When the entries in the table reach 0, i t i s assumed that the region i n question has been adequately pumped* Future.requests to either s t a r t or stop pumping w i l l disconnect the waste vacuum from a volume that has been pumped upon for a sat i s f a c t o r y length of time. , Pumping w i l l then begin on the highest p r i o r i t y area flagged as needing to be pumped. The pumping routines have proven to work extremely well* They provide the i n l e t system c o n t r o l l e r with a degree of sophist i c a t i o n that would be d i f f i c u l t to achieve without the use of a microprocessor. 121 V..THE AUTOMATED SYSTEM 5.1 Preliminary Remarks The automated system contains a great deal of distributed i n t e l l i g e n c e . The minicomputer system used by the isotopes group suffers from some disadvantages. Two of the most s i g n i f i c a n t disadvantages are: i t i s used to simultaneously control three mass spectrometers and three teletypes and yet possesses only 16 kilobytes of core.memory and i t i s programmable only in Interdata Assembler which l i m i t s the number of people that can program i t , as well as greatly increasing the e f f o r t and time necessary to develop new programs* , Both of these considerations resulted i n a system design that put the i n t e l l i g e n c e i n the mass spectrometer console.. This allows the operator to manually control the i n d i v i d u a l portions of the mass spectrometer i n any sequence he desires.. The only portion of the entire process in which the computer i s absolutely necessary i s i n the data acq u i s i t i o n and reduction stages.„ In addition to th i s e s s e n t i a l r o l e , i t normally, acts as the central c o n t r o l l e r for the system and i n i t i a t e s the control functions performed by the hardware and firmware devices i n the mass spectrometer's console.. This design therefore gives the operator s i g n i f i c a n t power and yet maximizes the system's f l e x i b i l i t y * . It also makes i t quite simple to 122 replace the Interdata minicomputer with a dedicated processor without losing a large amount of the e f f o r t presently i n the system*. A reasonable : replacement for the Interdata could well be the Motorola 68000 microprocessor that i s presently available.. The Motorola 68000 i s programmable in Fortran and Motorola assembler Language and could simulate the i n t e r f a c i n g conventions employed by Interdata minicomputers. 5.2 advantages of an automated Oxygen Isotope Mass Spectrometer The automated oxygen isotope mass spectrometer was designed using information and experience gained during the analysis of several hundred water samples used i n an M.Sc.,project., During t h i s time several problem areas were i d e n t i f i e d and t h i s information proved to be quite useful. I t was concluded that the single. most important consideration in the mass spectrometric analysis of carbon dioxide samples i s the consistency with which the operator treated each sample. The method of analysis i s such that many poten t i a l sources of error have an i n s i g n i f i c a n t e f f e c t on the measured DEL values. a s i g n i f i c a n t advantage of an automated mass spectrometer i s the consistency with which analyses are performed. For at least the l a s t decade the mass spectrometry laboratory at the University of B r i t i s h Columbia has 123 experienced great d i f f i c u l t y i n the area of continuity of laboratory personnel. The departure of a technician resulted i n the loss of much of the expertise required for the f a c i l i t y to function i n an acceptable: manner.. An automated f a c i l i t y greatly reduces the amount of expertise required to maintain a functioning f a c i l i t y . . Automation also increases the continuity of techniques employed i n the analysis of carbon dioxide samples.. I t i s true that the complexity of the instrumentation has increased s i g n i f i c a n t l y but the department has extremely capable electronics personnel to provide technical support.. The most obvious advantage of t h i s system i s the speed with which analyses are performed., At the time of t h i s thesis writing the system had been demonstrated to analyze carbon dioxide samples at a rate of more than 5 samples per hour.. After the i n i t i a l optimization of the mass spectrometer source potentials, the analysis of 34 carbon dioxide samples required less than 5 minutes of operator time f o r the periodic mounting of samples on the. i n l e t system. In the present configuration the operator i s ess e n t i a l l y free to perform the e q u i l i b r a t i o n of water samples with carbon dioxide and as such, the system has d r a s t i c a l l y increased the rate at which samples can be analyzed.. At the present time, plans are well underway for the construction of an automated sample preparation l i n e to work i n conjunction with the:automated mass spectrometer. In f a c t , the i n l e t system c o n t r o l l e r i s now capable of 124 direct interface with such a sample preparation l i n e which w i l l complete the automation of the f a c i l i t y * . An automated sample preparation l i n e i s now available from Micromass as item number MM5020. The addition of such a l i n e would be expedient (but expensive) and would result i n a completely automated f a c i l i t y with extremely high throughput and more than adequate precision. The degree of automation present i n t h i s mass spectrometer i s responsible for a combination of long term s t a b i l i t y and system v e r s a t i l i t y that i s extremely desirable. The technical innovations have produced a mass spectrometer with e s s e n t i a l l y unlimited long term s t a b i l i t y and precisions of better than 0.04°/oo.. The f l e x i b i l i t y of the system was demonstrated by the ease with which software modifications were made to convert the oxygen isotope mass spectrometer to a spectrometer also capable of measuring the carbon isotope r a t i o * . This w i l l be discussed i n more d e t a i l in a following section* _ 5.3 System Description The automated oxygen isotope mass spectrometer was designed with the preceding considerations used as guidelines. Figure 5.1 i s a block diagram of the f i n a l system., Of the 23 in d i v i d u a l components represented in the block diagram only 3 are remnants of the previous f a c i l i t y . A l l d i g i t a l components were designed f o r t h i s thesis FIGURE 5.1 BLOCK DIAGRAM OF THE AUTOMATED OXYGEN ISOTOPE' MASS SPECTROMETER The following block diagram depicts how the various components in the mass spectrometer system are interconnected. The arrows indicate the dir e c t i o n of information transfer. The dashed l i n e indicates either motion of carbon dioxide gas or carbon dioxide ions within the mass spectrometer.. The system i s conveniently subdivided into the f i v e categories shown at the extreme l e f t of the figure.. INTERDATA MINI COMPUTER RATIOMETRIC DIGITAL-VOLTMETER ^MEASURING [ BUT' WTH. SYSTEM J FILTER COLLECTOR • ~ l l i i L ROUGHING VALVE 1 ROUGHING PUMP 16 CHANNEL A / D CONVERTER BASELINE COMPENSATORS! BELL GAUSSMETER DIGITAL MAGNET CONTROLLER MAGNET CURRENT SUPPLY ANALYZER TOEPLER PUMP CONTROLLER, F ILAMENT EMISSION SUPPLY TOEPLER] 1 N L E T PUMP 1 SYSTEM GENERAL I/O MODULE INLET SYSTEM SUPPLY HIGH VOLTAGE SUPPLY SOURCE ION PUMP I. P. CONTROLLER I 127 project, with the. exception of the ratiometric d i g i t a l voltmeter which was purchased. With the exception of the filament supply a l l of the analog c i r c u i t r y was extensively modified for the i n s t a l l a t i o n of the automated system. Improvements i n the analog c i r c u i t r y were responsible for an increase in s t a b i l i t y from 5 minutes to 7 hours* With the exception of the magnetic analyzer i t s e l f , a l l components of the actual mass spectrometer are new to t h i s f a c i l i t y . The vacuum system i s capable of producing a pressure of 10~9 t o r r at the throat of the ion pump and of 5X10 - 8 torr at the ion gauge located near the source*_ These vacuums surpass any others obtained in our laboratory and are comparable to those obtained at other laboratories doing s i m i l a r analyses. ,. The block diagram describing the system shows how the various components are interconnected. In the automated mode of operation the minicomputer acquires data from the measuring system through the ratiometric d i g i t a l voltmeter as described by Russell and ahern (1974). The data are reduced to DEL Values by an assembly language; subroutine developed for t h i s system* Performance of the mass spectrometer i s quantitatively monitored by the minicomputer by c a l c u l a t i n g the standard error of the mean of the measured DEL values as data i s acquired. The.minicomputer makes l o g i c a l decisions as to whether to continue an analysis, issue a change sides request to the i n l e t system c o n t r o l l e r , issue, a request for a new sample to the i n l e t 128 system c o n t r o l l e r , or to reanalyze the.sample presently i n the unknown side Toepler pump* I t performs these actions through the general input/output (GIO) module. The minicomputer also performs such actions as ion beam suppression by biasing the magnetic f i e l d using the d i g i t a l magnet c o n t r o l l e r (DMC) and zeroing of the baselines of the parametric amplifiers i n the measuring system by act i v a t i n g the baseline compensation c i r c u i t through the: GIO module* The: minicomputer also accomplishes ion beam centering using the DMC and the A/D converter. I t s role w i l l be discussed more f u l l y i n the next section. The i n l e t system c o n t r o l l e r (ISC) has d i r e c t control of a l l portions of the i n l e t system with the exception of the Toepler pumps. The ISC issues commands to the Toepler pump co n t r o l l e r (TPC) when i t i s necessary for the.Toepler pumps to perform some action. The TPC uses the Butterworth f i l t e r e d ion beam i n t e n s i t i e s as feedback for the operations i t performs. The filament emission supply (FES) and high voltage supply (HVS) provide the currents and potentials required by the various portions of the source.. These supplies are in t e r n a l l y regulated and function i n a stand-alone fashion. They are optimized manually at the beginning of an analysis day and have acceptable i n t e r n a l s t a b i l i t i e s so as not to degrade the operation of the mass spectrometer., The filament emission supply i s the weakest component i n the system and should be replaced with a more modern supply 129 b u i l t by Peter Michalow of the department's ele c t r o n i c s shop*, I t should be noted that the long term d r i f t of these supplies can be compensated f o r by the other automated components. . The B e l l gaussmeter, magnet current supply and the electromagnetic analyzer are arranged in such a way as to provide 1 part i n 54,000 f i e l d regulation.. The e f f e c t of long term d r i f t in either the magnetic f i e l d intensity or high voltage i s eliminated by using a software algorithm i n the minicomputer i n conjunction with other components i n the system. The components of the:mass spectrometer and the vacuum system have been discussed i n previous chapters and w i l l not be discussed here. The interconnections between the various electronic supplies and the source and c o l l e c t o r make use of inexpensive high vacuum feedthroughs described by Gerber and Post (1973).. The s t a i n l e s s s t e e l jacketed coaxial cable was s i l v e r soldered into a prepared baseplate inside: an RF furnace.. This allowed precise temperature control over the entire extent of the baseplate. Individual feedthroughs were made 6.25 inches long since i t was found that heating of the te f l o n i n s u l a t i o n caused the teflon to soften and flow* , Feedthroughs that were. shorter then 2 inches i n length were found to lose t h e i r high vacuum c a p a b i l i t i e s due to t h i s extrusion. We were able to place 17 high vacuum, high voltage feedthroughs i n a 4 inch by 1.25 inch rectangular area.. 130 The measuring system i s s t i l l e s s e n t i a l l y as described by Russell and Ahern (1974). The need for baseline compensation noted in that paper was solved by the d i g i t a l baseline compensation c i r c u i t s shown i n Figure 3.2.. With t h e i r addition the measuring system has been found to be s u f f i c i e n t l y stable for long periods of unattended operation. In the previous oxygen isotope f a c i l i t y the teletype was only used to output intermediate measurements., In the present system the teletype.is used for outputting the f i n a l DEL values as well as indicating exactly how each sample has been treated by the minicomputer.. Requests for new samples and reanalysis of carbon dioxide due to low precision are c l e a r l y documented on the teletype.„ The teletype now serves as the:main l i n k between the operator and the device support routine.in the Interdata minicomputer. In t h i s command mode values of constants can be altered, commands can be given to hardware peripherals, peak centering Can be i n i t i a t e d , baselines can be zeroed and the program can be:placed i n the GO mode of operation where i t serves as the central c o n t r o l l e r f or the.automated system. 5.4 The Role of the:Interdata Computer 131 5.4.1 On-line Data Reduction The performance of the oxygen isotope f a c i l i t y can only be evaluated by multiple.analyses of water samples.. The performance of the mass spectrometer can however be evaluated by calc u l a t i n g the standard error of the mean of the sequential DEL values obtained during the mass spectrometric analysis of any one sample of carbon dioxide. Poor mass spectrometer performance can be: overcome by obtaining more estimates of the DEL value for that sample. I t i s therefore reasonable to discuss the method of data reduction used i n this study. Ahern (1975a) discusses in d e t a i l a method of o f f - l i n e data reduction using a f a i r l y sophisticated method of f i t t i n g cubic splines to the phi values output b-y the ratiometric d i g i t a l voltmeter. This method gave very good resu l t s and error estimates obtained were consistent with errors indicated by the r e p r o d u c i b i l i t y of the results.. The i n s t a l l a t i o n of the automated system required that the:DEL values be calculated within the minicomputer where the use of cubic splines would be most d i f f i c u l t . . For t h i s reason the simpler method of l i n e a r . i n t e r p o l a t i o n shown i n Figure: 5.2 was devised.\ A dir e c t comparison of the DEL 132 FIGUBE 5.2 THE METHOD OF LINEAR INTEBPOLATION USED IN THE DATA SEDUCTION The following figure compares the method of determining DEL values using cubic splines with a simple l i n e a r i n t e r p o l a t i o n method.. The two methods have been shown to produce equivalent r e s u l t s to within 0.03°/oo., The data are actual phi values (see E u s s e l l and Ahern, 1974) obtained from a re a l analysis. The upper l i n e (circles) are values from the l e f t hand side of the i n l e t system and the lower l i n e (triangles) are values from the right hand side.. 134 values and error estimates calculated by the; two methods showed that the. average difference between DEL values calculated using cubic splines and DEL values calculated by the l i n e a r interpolation method was 0.006+0.028.. The average difference in estimates of the standard error of the mean of the two methods was -0.004±0.006. These r e s u l t s indicate that the l i n e a r i n t e r p o l a t i o n data reduction scheme i s as reasonable to use as the more complicated method described i n Ahern (1975a).. The on-line data reduction i s performed i n two steps. In the automated mode, the Interdata computer i s interrupted p e r i o d i c a l l y by the ratiometric d i g i t a l voltmeter and a phi value i s passed to the computer. These phi values are d i g i t a l l y f i l t e r e d by a scheme sim i l a r to that described by Blenkinsop (1972)., After 7 f i l t e r e d data points are acquired and t h e i r average calculated, a c a l l i s made to a subroutine programmed by t h i s writer* „ The subroutine ascertains whether the data point came from the right hand side or t h e . l e f t hand side of the i n l e t system by sensing whether valve 9 (see Figure 4. 1) i s open or closed.. The program therefore functions c o r r e c t l y regardless of whether the standard or unknown i s analyzed f i r s t . . After 3 f i l t e r e d phi values are obtained the DEL estimates can be calculated using eguation 5*1a or 5.1b DEL = 2V-1 "^n ~^n-2 x 1000 STANDARD [5.1a] n 2k6-2<j> 135 DEL = +(^n-2 ~ 2 < i r i - l x 1000 UNKNOWN [ 5. 1b ] n 2k6 -4 -<j> n n-z where n i s the number of f i l t e r e d data points.. The;standard deviation and standard error of the mean are calculated when two or more DEL estimates have been obtained. The. r e s u l t s of the calculations are output to the teletype.„ They are also stored in the minicomputer memory to be used l a t e r by the computer when i t i s functioning in i t s role as central c o n t r o l l e r for the system.. The switching correction discussed by Ahern (1975a) i s performed within the Interdata minicomputer.. The tank carbon dioxide correction, the correction to the l o c a l tap water standard, the. l i n e a r world correction and the correction to V-SMOW are s t i l l performed o f f - l i n e by the Amdahl computer operated by the University of B r i t i s h Columbia's Computing Centre.. 5.4.2 The Minicomputer as the Central Controller As indicated in Figure 5.1, the Interdata minicomputer occupies an important position i n the automated system.. As mentioned i n the previous section i t processes the.data on-line and also monitors the performance of the mass spectrometer by continuously estimating the precision of the calculated DEL values. .. The Interdata minicomputer also allows the operator to 136 place the computer i n a command mode devised by B.D..Russell.. In his i n i t i a l configuration, the operator could enter the date and certa i n constants needed for data reduction into the computer memory._ The command repertoire has been extended to include commands which: 1._ center an ion beam i n i t s Faraday cup 2... zero the baselines of the parametric amplifiers 3. _ a l t e r the magnetic f i e l d to jump to the side of the peak 4. j a l t e r the magnetic f i e l d to jump to the top of the peak 5.. issue a CHANGE SIDES command to the i n l e t system c o n t r o l l e r 6. _ issue a request for a new sample to the i n l e t system c o n t r o l l e r . . One can also issue, the GO command which places the minicomputer i n i t s running mode where i t controls the automated system without operator intervention. In t h i s mode the minicomputer acquires data from the: ratiometric d i g i t a l voltmeter and calculates DEL values and error estimates as described i n the previous section.. After each c a l l to the subroutine that calculates DEL values the software i n the minicomputer (see Appendix I) determines which of the following four conditions exists: 1. _ There are . less than 3 measured DEL values.. 2. , There are between 3 and 8 measured DEL values and the estimated error i s les s than 0.04 DEL. 137 3.,, There are between 3 and 8 measured DEL values but the estimated error i s greater than 0.04 DEL. 4.. There are greater than 8 measured DEL values., The condition that exists determines what future action the minicomputer w i l l undertake i n i t s role.as the ce n t r a l c o n t r o l l e r . I f i t determines that either condition 1 or condition 3 exists then a "C" w i l l be output to the teletype indicating that the analysis i s Continuing._ The detection of condition 2 means that an acceptable analysis has been performed on the unknown carbon dioxide and the next unknown i s needed. The minicomputer outputs "NEW SAMPLE" to the teletype, spaces the paper and begins the following sequence of steps: 1.. A request for a new sample i s issued to the i n l e t system controller.> 2.. The program waits u n t i l the i n l e t system c o n t r o l l e r indicates that the next sample i s ready f o r analysis. 3. . The magnetic f i e l d i s altered so that no ion beams enter the.two Faraday cups.. 4., Control pulses are given to the baseline compensators and the e f f e c t s of of f s e t s i n the parametric amplifiers are eliminated. The minicomputer waits u n t i l both compensators have finished t h e i r tasks before continuing. 5. , The magnetic f i e l d i s returned to i t s o r i g i n a l value. , 138 6._ F i n a l l y the minicomputer performs a peak centering procedure that insures that the mass 46 ion beam i s centered i n i t s Faraday cup. The analysis of the next sample then begins*. In the event that more than 8 DEL values have been measured and yet the precision i s s t i l l unacceptable (>0.Q4) the minicomputer outputs "RON AGAIN" to the teletype, and performs steps 3 through 6 in the previous l i s t . . In t h i s manner the system can e s s e n t i a l l y eliminate bad data points and insure that an unknown sample i s measured with a precision of 0.04 DEL or l a s s . . Multiple analyses of unknown samples are combined using the "best" weighting function described by Ahern (1975a).. The automated system has been shown to function extremely well. . The assembly language program that controls the automated system i s presented in Appendix I.. 5.4.3 The Peak Centering Algorithm Unlike most portions of the automated oxygen isotope system, the peak centering algorithm i s l a r g e l y a resu l t of software within the Interdata minicomputer. ., The inherent s t a b i l i t y and highly favorable performance: of the peak centering algorithm warrant a b r i e f discussion of the technique used*. The success of the technique has resulted i n e f f e c t i v e l y unlimited long term s t a b i l i t y . 139 The peak centering algorithm makes use of the d i g i t a l magnet c o n t r o l l e r and the multiplexed analog to d i g i t a l converter shown in Figure 5.1.. The peak centering algorithm was chosen after f i r s t considering several other techniques. Although t h i s algorithm i s not always as fast as a more sophisticated approach t r i e d e a r l i e r , i t was found to be inherently stable and so r e l i a b l e that i t w i l l successfully center upon a peak even i f the ion beam i s not entering the Faraday cup. The beam must only be within one peak width of the edge of the Faraday cup.. The peak centering i s performed on the mass 46 ion beam since i t i s the smaller ion beam and therefore more d i f f i c u l t to measure.. The basic p r i n c i p l e behind the algorithm i s that i f the ion beam i s centered i n the Faraday cup, then incrementing the magnetic f i e l d so that only a portion of the ion beam enters the Faraday cup, w i l l give the same peak height as when the magnetic f i e l d i s decremented by the same bias from i t s central value; The algorithm developed for t h i s system consists of 12 ind i v i d u a l steps. They are as follows: 1. , The magnetic f i e l d i s incremented by an appropriate bias to jump to the.high mass side of the peak.. 2.. The minicomputer delays about 2 s to allow the peak height to reach a steady state. . 3.. The mass 46 peak height i s measured 32 times by a 12 bi t A/D converter and the average i s stored i n the computer as V1. 140 4. „ The magnetic f i e l d i s decremented by twice the bias used i n step 1 to jump to the low mass side of the mass 46 peak. 5. _ The minicomputer delays about 2 s. 6. . The mass 46 peak height i s again measured 32 times and the average i s stored as V2 inside the computer;, 7.. The minicomputer calculates V2-V1 and (V2+V1)/2. 8.. I f the absolute value of the difference i n low and high mass peak heights i s le s s than approximately 42 mV, then the ion beam was adequately centered and the f i e l d i s incremented by the bias to return to i t s centered position.. The peak centering i s complete.. 9.. I f |V2 - V l j i s less than 187.5mV then the peak side (fine) of the d i g i t a l magnet c o n t r o l l e r i s selected. Otherwise the scan side (coarse) i s selected. 10. The selected side of the magnet c o n t r o l l e r i s altered by +1 or -1 b i t depending on whether V2-V1 i s positive or negative. The magnetic f i e l d i s altered so as to move toward the peak side that produced the larger of the measured voltages.. i l l . After a delay of about 660 ms, the peak height i s measured 32 times and averaged*. I f the sign of (V2-V1) and (Peak Height-(V2+V1)/2) are the same then control i s returned to step 10. 12.. The scan side i s decremented by the; bias to jump back to the peak center and control i s transferred 141 to step 1. . I f any requested change i n the peak side setting would cause an overflow or underflow, the peak centering algorithm detects t h i s condition, resets the peak side to midrange.and alters the scan side appropriately to roughly recenter the ion beam.. The peak centering algorithm i s then reentered to insure the ion beam i s cor r e c t l y positioned in the Faraday cup. In most cases the peak i s found to be nearly centered and the entire process takes about 4 s to complete.. For testing purposes i t was found that even when the mass spectrometer i s measuring a baseline this algorithm w i l l accomplish peak centering i n l e s s than 30 s., The quality of the centering f a r exceeds that which a s k i l l e d operator normally produces even when taking up to 5 minutes to accomplish the task. 5.6 The V e r s a t i l i t y of the System The automated mass spectrometer system was designed with the single purpose of measuring the oxygen isbtope r a t i o i n carbon dioxide.. Nevertheless the unusual amount of distributed i n t e l l i g e n c e within the in d i v i d u a l components has resulted i n a system that i s e a s i l y modified by simple changes in the software within the Interdata minicomputer. a radio carbon group at Simon Fraser University headed by Dr..Earl Nelson contacted us i n 1978 to determine the ; 142 f e a s i b i l i t y of measuring the: r a t i o of C 1 3 / C 1 2 in carbon dioxide samples.. The basic d i f f i c u l t y in measuring t h i s r a t i o i s that the dual beam c o l l e c t o r possesses geometry designed to measure masses 44 and 46 and the C 1 3 isotope appears at mass 45. Certainly the best solution to t h i s dilemma i s to redesign the c o l l e c t o r and measuring system to allow t r i p l e ion beam c o l l e c t i o n but t h i s would not be a t r i v i a l task.. i n a lternative solution was possible due to the inherent f l e x i b i l i t y of the automated system.. The a b i l i t y '•• to control the magnetic f i e l d i n t e n s i t y using the Interdata minicomputer, allows data to be acquired from the mass 44 ion beam entering the mass 44 Faraday cup.. After a l t e r i n g the magnetic f i e l d appropriately, data can then be co l l e c t e d from the mass 45 ion beam entering the mass 46 Faraday cup. The.ion beams can be measured by the: 12 b i t A/D converter module;, These converters possess i n s u f f i c i e n t resolution to corre c t l y measure the r a t i o of the two ion beams to 0.1°/oo with a single conversion.. This d i f f i c u l t y can be:overcome by multiple sampling of an ion beam's magnitude, assuming that the noise present on the output of the parametric amplifiers i s larger than the resolution of 2i5 mV possessed by the A/D converter. I f the noise i s completely random then the e f f e c t i v e precision with which an ion beam can be measured increases as the square root of the number of conversions performed. To obtain a precision equal to a 4 1/2 d i g i t converter, at least 24 conversions 143 with a 12 b i t converter must be performed.. I f systematic noise also e x i s t s on the analog signal then even more conversions are required. I t would seem reasonable to make 32 conversions to insure acceptable precision with the present system. The changes i n the software necessay to implement the technique, described i n t h i s section were performed by Dr. , R. D. Russell. . This reprogramming took less than 3 days, and shows the r e l a t i v e ease with which the system can be modified. A detailed knowledge of the operation of the in d i v i d u a l components was not needed to make the necessary modifications. The results of t h i s i n i t i a l project demonstrated the f e a s i b i l i t y of the technique. Results seemed to indicate that the precision was limited to 12 b i t s but t h i s was most l i k e l y due to the fact that Dr. Russell sampled the outputs of the parametric amplifiers a f t e r they had been Butterworth f i l t e r e d . . The amplitude of the noise was therefore too small for the technique suggested i n thi s section to apply. The experiment should c e r t a i n l y be performed again using the unf i l t e r e d outputs before any conclusions can be reached. Nevertheless the v e r s a t i l i t y of the system has been demonstrated. I t appears that the r a t i o of any two isotopes in most gas samples can be measured with t h i s automated system without making any hardware changes.. 144 5.7 The Performance- of the Automated System The performance of the mass spectrometer was evaluated by running samples that had previously been analyzed on the unautomated mass spectrometer. Ahern (1975a) reports the measured DEL: values obtained from a suite of 12 samples that had been c a r e f u l l y prepared from a mixture of two i s o t o p i c a l l y d i s s i m i l a r water samples. , These samples have served as the laboratory test of procedures for several years and i t seemed appropriate to use them as the i n i t i a l test of the system.„ The 12 samples were analyzed along with 22 samples of tank carbon dioxide- on a single day* The results of the analyses are given i n Table 5.1.. ACCEPTED 1 ERROR | MEASURED j ERROR VALUE | | VALUE | —+ H ~ - 2 6 . 0 4 | . 0 3 | - 2 6 . 0 * | . 0 4 ^ 2 4 . 0 2 | . 0 9 H - 2 3 . 8 7 | . 0 4 - 2 1 . 6 7 | . 16 | ^ 2 1 . 8 2 | . 0 3 - 1 9 . 4 6 | . 10 | - 1 9 . 4 3 | . 0 4 - 1 7 . 2 3 | . 10 I - 1 7 . 2 6 | . 0 3 - 1 5 . 0 0 | . 0 6 1 - 1 4 . 9 8 | . 0 4 - 1 2 . 6 7 J . 0 8 I - 1 2 . 6 5 | . 0 4 - 1 0 . 4 9 | . 11 1 - 1 0 . 6 0 | . 0 3 - 8 . 3 0 | . 0 5 I - 8 . 1 4 J . 0 3 - 6 . 1 9 | . 09 | - 6 . 3 8 | . 0 4 - 4 . 1 1 | . 1 1 I - 3 . 9 3 | . 0 3 - 1 . 5 8 | . 0 5 | - 1 . 4 7 | . 0 4 TABLE 5.1 ISOTOPIC VALUES OBTAINED FROM A SUITE OF LINEARLY VARYING MATER SAMPLES The average difference:between the entries i n Table 5.1 i s 0.0 96 which implies that the precision of the analyses i s 145 about 0.08°/do (Youden,1951). This calculation assumes that the entries i n the table are true duplicates.. This i s not the case as they were obtained from d i f f e r e n t mass spectrometers.. Nevertheless, the ca l c u l a t i o n indicates that the performance of the automated system i s guite acceptable since 0*07°/oo of the error obtained can be attributed to errors i n sample preparation (Ahern 1975a).. i — . ' r : _ - " ' 1 I I | DEL VALUES RELATIVE TO V-SMOW I -34. 68*. 0 4 \ -34.71*. 03 1 -34.71+.03 ! -34.71 + .04 -34. 68 + . 04 | -34.67+. 04 ! -34.66+.03 I -34.60+.03 -34.62*.03 1 -34.63*. 04 I -34.62+.04 I -34.64+.04 -34. 64+. 0 2 | -34.64+. 02 ! -34.66+.02 j -34.68+.04 -34.69+.04 | -34.68+. 04 I -34.69+.04 J -34. 63+.03 1 -34.66+. 03 ! -34.65+.04 I MEAN=-34.66°/oo STANDARD DEVIATION =0.03°/oo TABLE 5. 2 THE RESULTS OF THE AUTOMATED ANALYSIS OF 22 ISOTOPICALLY IDENTICAL C02 SAMPLES The res u l t s obtained by measuring 22 samples of i s o t o p i c a l l y equal carbon dioxide samples provide: a clear i n d i c a t i o n of the precision with which the automated mass spectrometer system can determine the i s o t o p i c r a t i o of oxygen i n carbon dioxide samples.. Table 5.2 summarizes the results of t h i s experiment. , The r e s u l t s presented i n Table 5.2 indicate that the 146 automated oxygen isotope mass spectrometer has a precision of 0.3.3°/oo. The analyses were performed with the mass spectrometer system displaying no d i f f i c u l t i e s . The analyses presented i n Table 5.1 and Table 5.2 were performed with no operator intervention other than sample mounting and communication of that fact to the i n l e t system c o n t r o l l e r . The analyses were performed i n 6.5 hours which i s a rate of more than 5 samples per hour. Only 3 of the ,3.4 samples were automatically reanalyzed ! i n the mass spectrometer due to a measured precision greater than 0.04°/oo after 5 DEL estimates were obtained. To demonstrate the performance of the system on actual samples, several samples of ice taken from a v e r t i c a l borehole on the Steele Glacier (see Appendix III) were reanalyzed.. At the time of the analyses, the mass spectrometer possessed several problems due to the fact that i t had not been used for a considerable:length of time. The magnet supply had acguired long term d r i f t and the mass spectrometer i n l e t leak had been damaged re s u l t i n g i n an unsteady flow rate.into themass spectrometer. As a r e s u l t of these d i f f i c u l t i e s the performance of the. mass spectrometer was d e f i n i t e l y below standard* . The Interdata minicomputer program was modified so that the cutoff c r i t e r i o n f o r an acceptable analysis was set to 0.!0°/oo instead of the usual 0.04°/oo. r The results of the analyses are presented i n Table 5.3. When the precision of the values in Table 5.3 was 147 calculated using the method described by Youden (1951) and assuming that the two l i s t s were actually duplicate analyses (this neglects the fact that the actual mass spectrometer system had changed) a precision of 0.16°/oo was obtained. This value c o r r e c t l y r e f l e c t s the. lower quality of the analyses obtained but nevertheless indicates there are no fundamental d i f f i c u l t i e s attributable to the automated mass spectrometer system.. I t i s believed that the lower precision calculated from the values i n Table 5.3 i s d i r e c t l y traceable to the unsteady gas flow rates into the mass spectrometer*. Replacement of the i n l e t leak should r e c t i f y t h i s problem. I t i s also worth mentioning that even though the magnet supply was d r i f t i n g considerably, the peak centering algorithm c o r r e c t l y compensated for t h i s d r i f t at the.end of each analysis*; No operator intervention was required. . 148 SAMPLE | MANUAL | AUTOMATED NAME | ANALYSIS | ANALYSIS HI -2 | -26.64*.10 | -26.84+.04 HI -4 | -26.65+.12 1 N. E. . ' HI -6 | -27i04 + .10 I -26.93+.07 Hi -8 | -27.04+.10 1 N. E. . H1 -10 | -2 7. 96 + i 16 | -27. 91+.07 H1 -12 | -27.23+.10 | -26. 99+. 0 9 H 1 -14 ( -27. 63 + . 16 1 N.E.. HI -16 | -29. 10 + .13 | -29.20+.07 HI -18 | -27.42+.13 | -27. 57+. 0 5 H1 -20 | -27.23+.12 1 -27.43+.0 8 HI -22 | -28.40+.08 1 N.Ri". HI -24 | r28.66+. 12 1 N. E. HI -26 | -27.23 + . 08 | -27. 46+. 0 8 HI -28 | -27. 16+.06 | -26.93+.10 HI -30 | -29. 74+. 15 | -29. 63*. 0 6 HI -3 1 | -29.21+.14 | -29.44+.0 9 H 1 -32. 1 | -29.48+.09 I -29. 24+. 08 HI -34 | -26.52+.11 | -26.76+.04 HI -36 | -27.37+.10 | -27.18+.03 | — « N.E.. Implies sample not reanalyzed TABLE 5.3 THE RESULTS OBTAINED FOR THE REANALYSIS OF SELECTED SAMPLES FROM THE STEELE GLACIER A mass spectrometer precision of 0.04°/oo and a reduction i n analysis time from 60 to about 11 minutes per sample c l e a r l y establishes the Caliber of the automated system. This analysis time:will be reduced even further when the automated i n l e t system i s moved closer to the mass spectrometer. Theoretical c a l c u l a t i o n s indicate that an observed transient during i n l e t system side changing w i l l be reduced by a factor of 4 due to a decrease i n the distance 149 between the mass spectrometer leak and the mass spectrometer source I believe the major contributions of t h i s t h e s i s project are: the development of an automated system with very high precision and large throughput; the: analysis of fo c a l point locations i n a dual beam mass spectrometer; the design of an extremely sophisticated i n l e t system; the development of a method by which a l l samples can be admitted through a single.mass spectrometer leak; and the development of automated peak equalization c i r c u i t r y using the Toepler pump c o n t r o l l e r . . The use of a new material in the construction of the gas source i s also a useful contribution*. The large amount of automation present i n the system and the . method of computer-mass spectrometer interaction are also very s i g n i f i c a n t . . This automated system should enable the mass spectrometry laboratory at the University of B r i t i s h Columbia to play an important role in the measurement of the oxygen isotope r a t i o in carbon dioxide samples. . 150 LIST OF WORKS CONSOLTED Ahern, T.K. , (1975a) An o 1 8 / 0 1 6 study of water flow i n natural snow*. M,. Sc. Thesis, University of B r i t i s h Columbia, 164 pp.. Ahern, T.K. (1975b) A physical model of i s o t o p i c f r a c t i o n a t i o n i n the growth of closed system pingoes. Unpublished report, University of B r i t i s h Columbia, 29 P. Ahern, T.„ Pritchard, B..and Harris, T.„ (1980) Oxygen isotope exchange between a hen's drinking water and the calcium carbonate of an eggshell* . Submitted to Nature. Baertschi, P. (1976) Absolute 1 8 0 content of standard mean ocean water.. Earth and Planetary Science Letters. 31, 34 1-344. Barnard, G.P. (1953) Modern Mass Spectrometry* The In s t i t u t e of Physics, London, 326 pp.. . Begbie, P.J., Beckinsale, E.D., Freeman, N.J. .and Rowell, R. E. (1972) A bakeable changeover valve for high precision mass spectrometrie comparison of the isotope composition of gases* Review of S c i e n t i f i c Instruments. 43, No. 10, 1454-1455., Blenkinsop, J., (1972) Computer assisted mass spectrometry and i t s application to rubidium*strontium geochronology*. Ph.D., Thesis, University of B r i t i s h Columbia, 109 pp* . Cartan, M.L. . (1937) Sur l a f o c a l i s a t i o n des faisceaux de p a r t i c l e s chargee par deviation c i r c u l a i r e en champ magnetigue transversal. Journal de Physique.. No.. 11, 453-470. Clarke, G.K.C. „and J a r v i s , G. T. (1976) Post-surge temperatures i n Steele Glacier, Yukon Terr i t o r y * Canada. Journal of Glaciology. . J.6, No.. 74, 261-268. , Coleman, M.L. and Gray, J. (1972) An adjustable gas source i n l e t system for an isotope mass spectrometer.. Review of S c i e n t i f i c Instruments.. 43, No.. 10, 1501*1503. Craig, H. (1961) Standard for reporting concentration of deuterium and oxygen-18 i n natural waters*„ Science. 133, 1833-1834.. 151 Duckworth, H. E. , (1958) Mass Spectroscopy,. Cambridge University Press, London, 206 pp.. Epstein^ S. (1953) A mass spectrometer for the measurement of small differences i n isotope abundance rat i o s * . " Mass Spectroscopy i n Physics Research. National Bureau of Standards Circular 522, 133. Epstein, S. , and Mayeda, T. , (1953) Variations of the 0 1 8 content of waters from natural sources. . Geochimica et Cosmochimica Acta. 4, 213-224., EXAR Integrated Systems Inc. (1975) XR-2240 Programmable t i mer/counter. Data Sheet*. 8 p. F a i r c h i l d Semiconductor (1977) CMOS data book. , Publication number 223-12-0002-087/10k.. 490 pp. F a i r c h i l d Semiconductor (1975) Low power Schottky and macrologic TTL data book. Publication number 2022-12-0003-045/100M.. 294 p.. Fink, A. (1974) B.Sc* Thesis, University of B r i t i s h Columbia, 73 pp., Gerber,S. and Post,D. (1973) Inexpensive: high vacuum feedthroughs*, The Review of S c i e n t i f i c Instruments. . 44 , No. 3, 341-342.. Glen, J.W. , (1958) The mechanical properties of i c e . Advances in Physics. . 7, 254-265. Gonfiantini, R. . (1977) Consultant's meeting on stable isotope standards and i n t e r c a l i b r a t i o n i n hydrology and geochemistry, September 1976.. International Atomic Energy Agency, Vienna, 14 p.. Gooze, M (1979) How a 16-bit microprocessor makes i t i n an 8 b i t world.. Electronics.. September 27, 1979, 122-125.. Halsted, R.E..and Nier, A.O* (1950) Gas flow through the mass spectrometer viscous leak. , Review of S c i e n t i f i c Instruments.. 2\t No. 12, 1019-1021.. Hartman, B. (1979) 16 b i t 68000 microprocessor camps on 32 b i t f r o n t i e r . . Electronics, October 11, 1979, 118-125.. Herzog, R. (1934) Ionen- und elektronenoptische zy l i n d e r l i n s e n und prismen.. Z e i t s c h r i f t fur Physik, 89, 44 7-473. 152 Honig, R.E..' (1945) Gas flow in the mass spectrometer. Journal of Applied Physics. = 16, 646-654. Interdata Incorporated (1969a) Reference manual, publication #2 9-004R02. 98 pp. Interdata Incorporated (1969b) Systems Interface manual, publication #29-003R02. „ 52 pp. Kistemaker, J. (1953) The. influence of f r a c t i o n i z i n g and v i s c o s i t y effects in mass spectrometer gas handling systems. Mass spectroscopy i n physics research. U.S. National Bureau of Standards C i r c u l a r 522.. 243-247.. Kollar, F. . (1960) The precise intercomparison of lead isotope r a t i o s . . Ph.D.. Thesis, University of B r i t i s h Columbia. , 107 pp. , Loveless, A.J. and Russell, R.D... (1969) A strong-focussing lens f o r mass spectrometer ion sources.. International Journal of Mass Spectrometry and Ion Physics. 3, 257-266. McKinney, C.R. ., McCrea, J. M., Epstein, S., Allen, H. A. .and Urey, H.C. (1950) Improvements i n mass spectrometers for the measurement of small differences i n isotope abundance ratios.. Review of S c i e n t i f i c Instruments. 21, 724-730,. Mi t c h e l l , D. (1971) An spectrometer*. M.Sc* Columbia, 64 pp. Moser, H., S i l v a , C., S t i c h l e r , W. and Stowhas, L. (1972) Measuring the isotope content i n p r e c i p i t a t i o n i n the Andes* . The role of snow and ice i n hydrology. Proceedings of the Banff symposia, September 1972.. Beauregard Press Limited.. _1, 14-23. _, Motorola Semiconductor Products (1978) MCM2716 2048x8bit UV erasable prom, Motorola Semiconductor Products data sheet 5 p. . Motorola Semiconductor Products (1975) M6800 microprocessor applications manual, 714 pp.. Motorola Semiconductor Products (1975) M6800 microprocessor programming manual, 303pp.. Motorola Semiconductor Products (1979) MC6802 microprocessor with clock and ram, Motorola Semiconductor Products data sheet #AD1-436-RL . 28 p. on-line computer assisted mass Thesis, University of B r i t i s h 153 Motorola Semiconductor Products (1978) MC6821 peripheral interface adapter. Motorola Semiconductor Products data sheet # DS9435R1. 12 p. , National Semiconductor (1977) Pressure transducer handbook. 13 1 pp. . Nier, A.O. , (1940) A mass spectrometer for routine isotope abundance measurements.. Review of S c i e n t i f i c Instruments., V\, 212-2 16.. Nier, A.O. _ (1947) A mass spectrometer for isotope and gas analysis. Review of S c i e n t i f i c Instruments. 1.8, No. 6, 398-419. Nier, A.O. , Ney, £. P. and Inghram, M. G. (1947) A n u l l method for the comparison of two ion currents i n a mass spectrometer* . Review of S c i e n t i f i c Instruments. 1.8, No. . 5, 294-297. . Nye, J.F. . (1951) The flow of g l a c i e r s and ice sheets as a problem in p l a s t i c i t y . Proceedings of the Royal Society. A207,554-572.. Nye, J.F. (1957) The d i s t r i b u t i o n of stress and velocity i n g l a c i e r s and ice-sheets* Proceedings of the Royal Society. A 2 39, 113- 132 Paterson, W.S.B. , (1969) The Physics of Glaciers* . Pergamon Press, London, 250 pp* . Precision Monolithics Incorporated (1979) Dac 100 data sheet.Publication number 0342 70KRD. . 15.15-15 . 1 6 . . Precision Monolithics Incorporated (1979) Op-07 data sheet. Publication number 0342 70KRD. 6. 24-6. 29.. Precision Monolithics Incorporated (1979) Cmp-02 data sheet.Publication number 0342 70KRD. . 8.7-8.12.. Eu s s e l l , R.D., Blenkinsop, J., Meldrum* R. P. . and M i t c h e l l , D.L. , (1971) On-line.computer assisted mass spectrometry for geological research. Mass Spectroscopy. 1.9, No. ( I , 19-36. Russell, R.D. and B e l l i s , E.J. (1971) Mass spectrometer power supplies using a s i l i c o n controlled A.C.. Switch.. Mass Spectroscopy-Original Papers* , 1.9, No. 1, 37-47. Russell, R.D., Slawson, tf.F. and Ahern, T. K. , (1974) F i n a l i a o / i 6 0 report to the.Polar Continental Shelf Project. 22 p. 154 Russell, R.D. and Ahern, T.K. (1974) Economical mass spectrometer ion current measurement with a commercial parametric amplifier. Review of S c i e n t i f i c Instruments.. 45, No. 11, 1467-1469. Russell, R.D. .and Ahern T. K. . (1977) Mass spectrometer sample l i n e d i f f u s i o n processes.. Mass Spectroscopy, Original-Papers. . 25, No. 3, 217-221.. Sharp, R.P., Epstein, S..and Vidziunas, I.. (1960) Oxygen isotope r a t i o s i n the Blue Glacier, Olympic Mountains, Washington. Journal of Geophysical Research. 65, No. 12, 404 3-4 05 9. Stacey, J.S.: (1962) A method of r a t i o recording for lead isotopes i n mass spectrometry. Ph.D.. Thesis, The University of B r i t i s h Columbia, 173 pp. Stanley, A. D. . (1969) Observations of the surge of Steele Glacier, Yukon T e r r i t o r y , Canada,. . Canadian Journal of Earth Sciences* 6,8 19-830.. Stephens, W.E. (1934) Magnetic refocussing of electron paths. Physical Review*, 45,513-518. , Teledyne Semiconductor (1977) Monolithic CMOS A/D Converters-8700 series.. Publication number DG-113-47-15M. 8 p. , West, K.E. (1972) H20* 8/H2Oi6 variations i n i c e and snow of mountainous regions of Canada.. Ph.D.. Thesis, University of Alberta 123 pp.. Whittles, B.L. (1960) Voltage c o e f f i c i e n t of Victoreen high-meg r e s i s t o r s . . Review of S c i e n t i f i c Instruments. . 31 No. 2, 208-209. . York, D. (1969) Least squares f i t t i n g of a straight l i n e with correlated errors*. Earth and Planetary Science Letters. 5, 320-324. Youdea, W.J.. (1951) S t a t i s t i c a l methods for chemists.. John Wiley and Sons, Inc., New York. 126 pp. , APPENDIX I INTER DATA ASSEMBLER PROSRAM The following l i s t i n g contains the assembly language source statements used to control the automated system*. The machine language i n s t r u c t i o n s are at the extreme l e f t of the o utput* Microfische copies of the Interdata program can be obtained by contacting the Department of Geophysics and Astronomy at the University of B r i t i s h Columbia. APPENDIX II MOTOROLA ASSEMBLER PROGRAM FOR THE INLET SYSTEM CONTROLLER Tie following l i s t i n g contains t i e Motorola assembly language source; statements used i n the i n l e t system c o n t r o l l e r . The machine language i n s t r u c t i o n s are at the; extreme l e f t of the o utput* Microfische copies of the Motorola program can be obtained by contacting the Department of Geophysics and Astronomy at the University of B r i t i s h Columbia. 157 APPENDIX II I ISOTOPIC EVIDENCE OF SURGES OF THE STEELE GLACIER, YUKON TERRITORY Oxygen isotope r a t i o s i n snow and ice samples often provide information about the physics governing the flow of g l a c i e r s and i c e sheets.. Due to the time scales one must deal with when studying g l a c i e r s , isotopic evidence provides a poweful experimental method of studying p a r t i c l e t r a j e c t o r i e s within a g l a c i e r . . Oxygen isotope ratios were used to study the flow of the Steele Glacier,* , The Steele Glacier i s a large valley g l a c i e r located approximately 70 km west of Burwash Landing, Yukon T e r r i t o r y . , The Steele i s i n a class of glaciers known to make sudden rapid advances, or surges*. The Steele Glacier i s thought to surge about every 75 years and most recently surged during a 2 to 3 year period sta r t i n g i n 1965. The pre-surge. terminus of the g l a c i e r was located approximately 29 kilometers from i t s main accumulation area on the slopes of Mount Steele. After the surge, ice had been displaced about 15 km down the valley. . Dr. G.K.C. Clarke of the Department of Geophysics and Astronomy at the University of B r i t i s h Columbia has studied the Steele Glacier for many years and has spent considerable time measuring the v e r t i c a l temperature.distributions within the glacier (Clarke and Jarvis (1976)).. These, measurements began i n 1972 and concluded in 1975 with the d r i l l i n g of two shallow thermal d r i l l holes.. In 1975 we began a study of the d i s t r i b u t i o n of oxygen isotopes in the Steele Glacier. 158 As part of t h i s study, the 1975 thermal d r i l l holes were "bailed" at 1 or 2 m i n t e r v a l s and samples of the water were colle c t e d for l a t e r analysis. In addition to t h i s v e r t i c a l sampling of the Steele Glacier, 74 surface samples were colle c t e d along 7 transverse l i n e s across the surface of the g l a c i e r from just below the f i r n l i n e to just above the point where Hazard Creek i n t e r s e c t s the_Steele :Glacier. Four of these transverse.lines were located within 3 km of the v e r t i c a l thermal d r i l l hole (see Figure A3.1). The samples were analyzed by R.D..Ahern i n the mass spectrometry laboratory at the University of B r i t i s h Columbia. , This served as the f i r s t r e a l test of the system designed by t h i s writer and the results were.obtained with a measured precision of 0.03°/oo. . The most inter e s t i n g r e s u l t s came from the samples collected from the 36 m thermal d r i l l hole.. The results are presented i n Figure A3.2.. The most notable feature of the iso t o p i c r a t i o s obtained from the thermal d r i l l hole i s the periodic "sawtooth" pattern exhibited i n the measurements.. As West (1972) points out, normal theories of deformation of g l a c i e r s in equilibrium predict a monotonic decrease i n the. o * 8 / 0 1 6 r a t i o i n a v e r t i c a l borehole. (see Figure A3.3).. This corresponds to the decrease i n the 0 1 8 / 0 1 6 r a t i o one encounters as one proceeds up g l a c i e r above the equilibrium l i n e (West, (1972) ; Sharp et a l . , (I960)).. This decrease i s most l i k e l y due to changes in the temperature at which the snow f a l l i n g in the FIGURE A3. 1 MAP OF THE STEELE GLACIER SHOWING SAMPLE LOCATIONS The following map indicates the locations from which surface ice samples were collected on the Steele Glacier*, The f i r n l i n e i s located about one kilometer west of the uppermost traverse. _ The v e r t i c a l borehole discussed i n t h i s section i s indicated by the dot within the c i r c l e * The Rusty Glacier from which West coll e c t e d his samples i s located just north of the Hodgson Glacier and i s indicated by the c i r c l e containing the cross. . 160 16 1 FIGORE A3.2 THE MEASURED OXYGEN ISOTOPE VARIATION AS A FUNCTION OF DEPTH TAKEN FROM THERMAL DRILL HOLE NUMBER ONE The measured DEL values, r e l a t i v e to Vienna SMOW, are displayed in the following graph.. The error bars on the v e r t i c a l axis represent the errors involved in the measurement of the.isotopic r a t i o while the errors on the horizontal axis represent the uncertainty i n the depth from which the samples were collected., 162 163 FIGURE A3.3 ISOTOPIC PATTERNS EXPECTED FROM A GLACIER IN EQUILIBRIUM This figure. summarizes the usual isotopic d i s t r i b u t i o n within a g l a c i e r that i s in equilibrium with i t s environment,. As w i l l be shown i n th i s section, t h i s d i s t r i b u t i o n i s not to be expected i n a surge-type g l a c i e r . , F I R N L I N E A C C U M U L A T I O N A R E A A B L A T I O N A R E A SO"3 I N V E R T I C A L B O R E H O L E S O 1 9 ON G L A C I E R S U R F A C E 1 6 5 accumulation basins of the Steele Glacier was formed. It i s also possible that the observed decrease i n the o 1 8 / 0 * 6 r a t i o i s a reservoir e f f e c t (Moser et a l . (1972)). West (1972) found that the nearby Rusty Glacier possesses a DEL gradient of -0.4°/oo per 100 m of elevation change. Although t h i s value: i s s l i g h t l y higher than that found i n other locations i t i s a reasonable value to use for the Steele Glacier due to i t s proximity to Rusty Glacier.. The ragged sawtooth pattern shown in Figure A3.2 i s believed to be the r e s u l t of the surging of the Steele Glacier. , This i s the f i r s t time that g l a c i e r surging has been detected i s o t o p i c a l l y and t h i s finding may have important implications for the detection and understanding of rapid advances of ice sheets should they have occurred i n the past. The physics governing the nonsteady flow of g l a c i e r s i s extremely involved and i t i s f e l t that the sampling technique used to obtain the samples from the Steele Glacier does not warrant a detailed analysis of the;glacier flow. The sawtooth pattern displayed by the isotopic r a t i o s i n the Steele Glacier can nevertheless be explained i n a semi-guantitative manner. Paterson (1969) provides an excellent summary of present t h e o r e t i c a l treatments of g l a c i e r flow. Most " r e a l i s t i c " explanations of p a r t i c l e : t r a j e c t o r i e s i n g l a c i e r s must s t a r t with the flow law of i c e . J.W. Glen (1958) indicates that the flow law of ice i s given 166 by an expression of the form E = A T n [A3. 1 ] •'. xy xy where Exy i s the s t r a i n rate r e s u l t i n g from the shear stress Txy.. The constant "n" i s independent of temperature and has the approximate value of 3 and "A" i s a temperature dependent constant. Nye (1957) discusses how to generalize Glen's Flow Law to r e a l g l a c i e r s * The form of equation A3.1 remains the same but with an e f f e c t i v e shear stress and an e f f e c t i v e strain rate suggested by Nye replacing the appropriate terms.. The generalized flow law given by A3.1 plays a c r u c i a l r o l e i n the determination of v e l o c i t i e s and p a r t i c l e t r a j e c t o r i e s i n glaciers* To study the d i s t r i b u t i o n of oxygen isotopes i n the Steele Glacier, a very s i m p l i f i e d physical model was developed.. The Steele Glacier was assumed to be a rectangular block of ice that i n i t i a l l y had the i s o t o p i c d i s t r i b u t i o n shown in Figure A3.4.. The gradient for the DEL values corresponds to a change of 1°/oo in 1 km.. Assuming a -0.5°/oo change i n the i s o t o p i c r a t i o per 100 m elevation change.. This corresponds to an elevation change of 200 m/km which i s quite reasonable for the upper accumulation zone of the Steele Glacier (Stanley, (1969)).. The simple model assumes that the ice i s accumulated without any i c e deformation u n t i l the pre-surge thickness of the g l a c i e r i s reached.. The i n i t i a l " i s o d e l " l i n e s ( l i n e s FIGURE A3.4 THE INITIAL POSITIONS OF TWO ISODEL LINES IN THE STEELE GLACIER The i n i t i a l model used in t h i s study assumes that the Steele Glacier simply accumulates'ice u n t i l the f i r s t surge takes place. The isodel l i n e s are shown i n the following figure, assuming that the i s o t o p i c r a t i o of the p r e c i p i t a t i o n at any given l o c a t i o n on the g l a c i e r i s . constant. . The distances are measured from the.headwali of the g l a c i e r . Later models assume that the isodel l i n e s are not v e r t i c a l but rather t i l t e d , assuming that the isotopic r a t i o of the snow f a l l i n g on the Steele Glacier i s determined by the: surface elevation of the glacier* , 168 169 of constant isotopic ratio) are:perpendicular to the surface of the:glacier which corresponds to a si t u a t i o n where the snowfall at a given distance up the g l a c i e r has a constant i s o t o p i c r a t i o . . The laws governing the deformation of ice are derived by Nye (1957)., For the simple model shown i n Figure A3.4 and with the assumption that the g l a c i e r i s s u f f i c i e n t l y wide so that the ef f e c t s of the valley walls are n e g l i g i b l e , Nye shows that the horizontal and v e r t i c a l v e l o c i t i e s are given by u=Uo+rx+f(y) [ A 3 . 2 ] and v=r (h-y) [ A 3 . 3 ] where Oo i s the basal s l i d i n g v e l o c i t y , r i s the lon g i t u d i n a l s t r a i n rate, h i s the thickness of the g l a c i e r and f(y) i s given by f ( y ) = -2Apgsin(a) fYQ yT 2 dy: n=3 [ A 3 . 4 ] where T 2 i s given by T 6 - T 4 { p g y s i n ( a ) } 2 - ( r / A ) 2 =0 [ A 3 . 5 ] where p i s the average density, g i s the acceleration due to gravity and a i s the surface slope of the g l a c i e r . Equation A 3 . 5 i s a cubic equation i n T 2 and as such can be solved a n a l y t i c a l l y . . The cubic determinant can be shown to be D= r 2 / A 2 [ (pgysin ( a)) 6/27*r*/4] \ [ A 3 . 6] and since a l l variables are r e a l , D i s always >0 and i t 170 can be:deduced that equation 43.5 has 1 real and 2 imaginary s o l u t i ons. The r e a l solution i s given by T2 = ^ p g y s i n ( a ) j 2 + ^ { p S y s i n ( a ) } 6 _ i_ + V D ' ] 1 / 3 [A3.7] '{pgysin(a)} 6 _ x_ _ /- ] 1/3 27 2 -• I z / 2A Z " -1 + Since. A3.7 determines T 2 a n a l y t i c a l l y we can e a s i l y calculate f.(y) from equation A3.4 and the horizontal and v e r t i c a l v e l o c i t y solutions are complete. . The d i s t o r t i o n of the iso d e l l i n e after a surge can be determined by applying equations A3.2 and A3.3 to i n d i v i d u a l points on the is o d e l line,. The surge was assumed to continue u n t i l the.surface component of the:isodel l i n e had extended 2 km down gl a c i e r . . This extension was chosen so that the amplitudes displayed on Figure A3.2 could be explained. Greater extensions would produce larger amplitudes in the observed i s o t o p i c variations. . The extension used i s consistent with observations made by Stanley (1969) for the portion of the Steele Glacier near the thermal d r i l l hole i n question.. Figure A3.5 shows the d i s t o r t i o n in the isodel l i n e s after the i n i t i a l model i s allowed to surge.a single time. Since the horizontal velocity i s greatest at the surface of the glacier the position of the isodel l i n e on the surface moves the farthest down glacier.„ After the surge the: model assumes that the g l a c i e r FIGHEE A3.5 THE SHAPE OF THE ISODEL LINES AFTEE A SINGLE SURGE For the simple model depicted i n Figure A3.4, a single surge of the Steele Glacier d i s t o r t s the v e r t i c a l i s o d e l l i n e s into the shape shown i n the following figure. f This d i s t o r t i o n follows d i r e c t l y from the Nye model of gl a c i e r deformation. . The two l i n e s correspond to isodel l i n e s of -29°/oo and -27°/ooi ZL I 173 again accumulates snow with the i s o d e l l i n e i n the newly accumulated ice located at the previous position as depicted in Figure A3.5. . The.glacier i s allowed to surge again and the result of multiple .surges i s that the i s o d e l l i n e s do indeed take on a sawtooth c h a r a c t e r i s t i c * The above model was modified to accommodate a more r e a l i s t i c assumption about the nature of the i s o d e l l i n e s i n the newly accumulated snow.. Paterson (1969) points out that gl a c i e r s do a f f e c t the l o c a l climate.. A surge, of a magnitude displayed by the Steele Glacier might well cause changes in the weather patterns within the Steele Creek Valley.. The assumption that snow of a constant i s o t o p i c r a t i o f a l l s at the same geographic location may well be i n v a l i d . . I t i s not unreasonable to make theiassumption that i t is the elevation of the surface of the g l a c i e r that determines the isotopic r a t i o of the snowfall. With t h i s modification the i s o d e l l i n e s were found to d i s t o r t into a pattern shown in Figure A3.6 after two surges.. After ten surges the.pattern shown i n Figure A3.7 was obtained. The sawtooth character i s c l e a r l y evident i n Figure A3.7 i t would seem that the variations found experimentally can be explained by combining a l o g i c a l model of isotope d i s t r i b u t i o n in the accumulation zone with an accepted model of g l a c i e r flow (Nye, (1957)). There remains one d i f f i c u l t y with the isotopic patterns displayed i n Figure:A3.7- The model produced four sawtooth variations i n a v e r t i c a l distance of 75 m which i s within a factor of two 174 FIGURE A3.6 ISODEL PATTERN FORMED AFTER TWO SURGES FOR AN INITIALLY NON-VERTICAL ISODEL LINE After a single surge of the g l a c i e r , t h i s model assumes that no motion of the glacier takes place u n t i l the next surge., The glacier simply accumulates snow*. The:discontinuity i n the isodel l i n e occurs because the i n i t i a l i s o d el l i n e s are displaced down g l a c i e r * , The is o d e l l i n e i n the newly accumulated snow i s positioned up gla c i e r since i t s position i s determined by the:. surface elevation of the gla c i e r . . i FIGURE A3.7 ISODEL PATTERN FORMED AFTER TEN SURGES FOR AN INITIALLY NON-VERTICAL ISODEL LINE Neglecting the longitudinal s t r a i n rate term i n the equation for the horizontal velocity within the g l a c i e r , yields a d e f i n i t e sawtooth pattern in the i s o d e l l i n e s * . The:portion of the curves below about 50 m i s a remnant of the i n i t i a l isotopic d i s t r i b u t i o n assumed and should be ignored. 178 of the number of cycles displayed in Figure.A3.2.. However, the horizontal scale of the variations i s 4 km* A v e r t i c a l borehole would intercept at most three cycles of the pattern and more r e a l i s t i c a l l y two cycles. The d i f f i c u l t y with the model i s that the lower portions of the isodel l i n e s are displaced too far down glac i e r to produce acceptable v e r t i c a l variations. The model was modified s l i g h t l y by assuming that longitudinal strain rate i s not constant but changes from a p o s i t i v e value i n the upper portions of the glacier to a negative value down g l a c i e r . . Examination of equations A3.4 and A3.7 indicates that f (y) i s not dependent upon the sign of the l o n g i t u d i n a l s t r a i n rate, but only upon i t s magnitude. Examination of equations A3.2 and A3.3 indicates that horizontal v e l o c i t i e s w i l l decrease down g l a c i e r when r i s negative due to the rx term and v e r t i c a l v e l o c i t i e s w i l l be toward the g l a c i e r surface. This w i l l tend to bring the sawtooth pattern i n Figure A3.7 closer to the surface as well as making the i s o d e l l i n e s more v e r t i c a l since horizontal v e l o c i t i e s w i l l decrease as one proceeds down glaci e r . . The above modifications were made to the:model and the r e s u l t s of the computations are presented in Figure A3.8. The most s i g n i f i c a n t difference between the r e s u l t s of t h i s f i n a l model and the model r e s u l t s obtained considering a smaller l o n g i t u d i n a l s t r a i n rate i s that the isodel l i n e s are much more v e r t i c a l . . In fact the model produces f i v e 179 FIGURE A3.8 POSITION OF TWO ISODEL LINES OBTAINED AFTER TEN SURGES OF A GLACIER EXHIBITING EXTENDING FLOW IN THE VICINITY OF THE BOREHOLE IN QUESTION The correct handling of the longitudinal s t r a i n rate in the equations governing ice deformation r e s u l t s i n the isodel l i n e s becoming more nearly vertical;. , The s t r a i n rate used to produce the vari a t i o n s , seen i n the following figure, was 0.1 per year.. The amplitude and period of the variations shown i n t h i s figure correctly predict the shape of the sawtooth pattern displayed i n Figure A3. 2. HEIGHT ABOVE BED O o 75.0 150.0 225.0 300.0 375.0 u> i I I I 1 1  D 08 1 18 1 surge cycles i n 1 km of horizontal distance* _ The models presented in t h i s section demonstrate that the sawtooth pattern observed in the isotopic ratios taken from a v e r t i c a l borehole can be explained by combining an accepted theory of glacier flow with a reasonable model of isotope d i s t r i b u t i o n i n the accumulation zone.. The sawtooth pattern i s produced by ice deformation during the surges of the Steele Glacier and i c e deformation during the quiescent period i s neglected.„ The e f f e c t s of the accumulation-ablation function are neglected i n the model with the exception that t h i s function i s important i n determining the longitudinal s t r a i n rates at any point within the glacier* 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0052992/manifest

Comment

Related Items