ISSN:0003-2654    

DOI:10.1039/AN98106FX009

出版商:RSC    

年代:1981

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98106BX011

出版商:RSC    

年代:1981

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98106FP029

出版商:RSC    

年代:1981

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98106BP035

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

MARCH 1981 The Analyst Vol. 106 No. 1260 The Impact of Microprocessors on Analytical Instrumentation A Review D. Betteridge and T. B. Goad Chemistry Department University College of Swansea Singleton Park Swansea SA 2 8PP Summary of Contents Introduction Development of microcomputers Applications of microcomputers to analytical instrumentation History of the development of an automated titrator Laboratory prototype Control laboratory model 0 ther approaches Conclusion Glossary Keywords Review ; microprocessors ; analytical instrumentation ; automated titrator Introduction In the last 6 years microprocessors* have progressed from being a novel exciting develop-ment of potential interest to analytical chemists into part of the everyday experience of the populace.Technical advances are still being made at a phenomenal rate but it is possible now to survey the scene from the point of view of the analytical chemist to assess the impact of microprocessors on analytical instrumentation to make some sense of the technical aspects and the jargon used to describe them and to discuss approaches to the application of the microcomputer. It is therefore practical and economic to use them in small-scale inst rumentation such as titrators polaro-graphs and balances. Two years ago it was less profitable to use microcomputers for large instruments such as mass spectrometers because of their relatively small memory and speed of operation. However improvements have been so rapid that these considerations are now less relevant.In analytical instrumentation they can fulfil a control function which may range from a simple replacement of mechanical or electromechanical parts with a chip to execution of a program of sequential operations that comprise the total analytical deter-mination. In addition they may be used with advantage for data acquisition and for data processing great advances having been made in respect of the latter over the last year. These aspects are discussed in detail in this paper but first the development of micro-processors and the current availability of microcomputers will be reviewed. The microprocessor-based microcomputer is compact cheap and robust. Development of Microcomputers Until the 1940s the principal devices for digital computation remained the traditional fingers and abacus.Babbage’s calculating engine provides a reminder of the practical diffi-culties of constructing a mechanical computer. During the 1940s several valve-based electronic digital computers were constructed and these provided a sound base for the developments that have led to the microprocessor. * Words that are italicised are defined in the Glossary at the end of this paper. 25 258 BETTERIDGE AND GOAD THE IMPACT OF Analyst Vol. 106 The valve-based computer on account of the valves and associated circuitry was large, dissipated a lot of heat and had many wires and connections. As a consequence it was difficult to accommodate uncomfortable to operate and subject to frequent breakdown. It was also very expensive to construct and maintain. The replacement of the thermionic valve by the transistor in the late 1950s led to a reduction in size and consumption of power.The concept of putting electronic circuitry on boards and assembling these into large units greatly increased the reliability of the large unit and made it easier to maintain. At this point the large computer was dominant the opportunities for reducing the size of the computer having been traded off against an increase in computing capacity. The computer was being widely used but in a hierarchical manner. The experimenter had to link his apparatus to the company’s mainframe computer possibly via a minicomputer and suffer all the inconveniences of time sharing. Further because of the inherent cost and its capacity it could only be justified for massive calculations and large-scale analytical opera-tions.In 1965 Digital Equipment Corporation introduced the PDP-8 minicomputer at a price of $18000 accessories such as printers and backing store being extras. However mini-computers because they were mainly associated with large instruments such as nuclear magnetic resonance (NMR) and mass spectrometers made only a small impact compared with the microcomputers that came later. Nevertheless between 1965 and 1975 28000 PDP-8s were installed. Thus in the late 1960s the basic computing rationale was a legacy of the valve computer, its large size and the magnitude of the breakthrough in computations. However the con-struction of a shoe-box sized computer with the calculative capacity of the early mainframe devices with minimal power consumption and maintenance problems and costing less than $5000 wrought a change in computer usage as profound in its long-term effect as the technological development that led to the microprocessor.In the mid-1960s the electronics industry which had successfully developed the concept of circuit boards and solid-state devices devised a process whereby integrated circuits could be etched on to a silicon chip. The resultant product can be made cheaply for a mass market, it has low power requirements and is robust and reliable. Then in 1971 a development equivalent in size and consequences to that of integrated circuits brought the computer well and truly into the laboratory. Intel Corporation a semiconductor manufacturer that supplied integrated circuits used in many minicomputers, devised the first microprocessor chip the Intel 4004.It was originally intended as a “controller” chip incorporating a complete central processing unit on a single chip. It required a number of other chips to make it into a complete computer control system and only four binary digits could be processed at a time. In the following year Intel brought out another microprocessor chip the 8008 which was faster and more powerful than the 4004; it had a 48 instruction set and could handle data in units of 8 binary digits. The 8008 became competition for those minicomputers that were dedicated to simpler tasks because although the cost of the chip was at first high it soon dropped to the $100 range. Intel’s next microprocessor chip was the 8080 again a great advancement over the previous genera-tion.However by this time other semiconductor manufacturers had started to produce microprocessor chips some offering new architecture designs and instruction sets others duplicating other manufacturers’ designs. Motorola Semiconductor’s MC6800 micro-processor chip and MOS Technology’s MCS6502 chip became two of the most popular competitors to the 8080. With an increase in the range of chips available came a decrease in price not only of the microprocessor chips but also of the support chips i.e. memories and input /out@ut devices. The first successful low-cost “home computer” was introduced by Micro Instrumentation and Telemetry System (MITS) in early 1975 and was called the Altair 8800. Based on the 8080 microprocessor chip it had a limited memory a control panel power supply and a complete packaging.It was an instant success selling for under $500. MITS soon followed the Altair 8800 with a 6800 based model. On the basis of the success by MITS other manu-facturers started to produce microcomputers some very similar in design to the Altair even to the extent of using the same bus structure for their plug-in boards. Others were totally different designs based on the 8080,6800,6502 or other microprocessors available. Currently there are many products being offered ranging from the bare essentials of the kits to th March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 269 complete microcomputer system with accessories. Table I lists those available at the time of going to press.A basic system consisting of microprocessor memory visual display unit and keyboard can be obtained for as little as L500 with accessories being added as the budget allows. For around A2500 a complete computer system can be obtained for use as a business system or in a control situation which is equivalent to and in some instances surpasses the typical minicomputers of 8 years ago. With the advent of the microcomputer the analyst in possession of a willingness to learn basic programming technique and a modicum of patience can develop many useful programs without recourse to an outsider. Applications of Microprocessors to Analytical Instrumentation Table I1 surveys many of the past and current chemical applications that involve micro-processors.In general the mode of operation for a microprocessor can be split into two categories firstly as a “controller” and secondly to accumulate and process data. The choice of which mode to use reflects individual preferences; some of the applications contained in Table I1 involve both modes others only one. The subject of microprocessors in analytical instrumentation is such a rapidly growing area that Table I1 can in no way be considered complete but as many references as possible have been included. As guidelines to the details in Table 11 we note the following general trends in the analytical applications of microprocessors : 1. Spectroscopy. A number of ultraviolet - visible and infrared instruments are appearing in which the settings are altered and optimised by microprocessors and in which the peaks are identified and quantified.Similar developments are being made in other branches of spectroscopy and the matching of image-sensing devices with micro-computers will make a big impact over the next decade. 2. Chromatography. A number of commercially available chromatographic integrators make use of a program for peak detection and measurement. The qualitative and quantitative results are conveniently displayed and doubtful assignments are indicated so that the operator may check the results and use his own judgement. The control of the chromatograph may also be carried out by microprocessors. 3. Electroanalytical chemistry. Output from electroanalytical determinations such as polarography or potentiometry with ion-selective electrodes can be readily processed by microcomputers with resulting improvements in quality.Most manufacturers and many authors discuss only very briefly if at all the options open to the chemist who wishes to apply microprocessors to his instrumentation and give little account of the practical difficulties that may be encountered. These practicalities are well illustrated by considering the case history of the design and construction of a micro-processor-controlled automatic titrator. History of the Development of an Automated Titrator Between 1975 and 1979 several groups devised microcomputer-controlled titrator~.~J3~3~~ 35*41*53972 Because of differences in their approach and the available state of microprocessor development the instruments proposed differed in several respects.The first were carried out simultaneously but independently in our own laboratories7 and in those of Radiometer8 and Mettler. Ours was basically a collaboration between an electrical engineer (E. L. Dagless) an industrial chemist (D. R. Deans) and an academic chemist (D. Betteridge) with a research student (P. David) working in whichever laboratory was appro-priate at the time. The problem as posed was to see whether a mixture of organic acids and boric acid from the ICI nylon process plant could be titrated sequentially without addition of mannitol and with an accuracy of 0.5% or better. The existing procedure was manual and consisted in first titrating potentiometrically the organic acids and then the boric acid after addition of mannitol to a pre-determined potential.Laboratory Prototype microprocessor. The first stage was to find a solution to the chemical problem given the capabilities of the Three possibilities were considered (i) to mimic a conventional automati 260 TABLE I-MICROCOMPUTE Manufacturers of the various microprocesoo Hardware description of basic system Micro- Minimum Interfaces and Type of backing Manufacturer Unit Configuration processor+ memory Visual display Keyboard 1/0 ports memory Other Acorn Acorn Micro- Consists of 8 digit LED 25 key 16 way I/O Computers computer two Eurocards display VDU (16 hex with l28B of System I -MPU card 6502 1KB RAM interface on 9 conttol) RAM on pro-and keyboard extra card keyboard cessor card. card. Buy in kit form or interface on assembled CUTS cassette keyboard card Apple I1 Home 6502 16KB RAM Extra Keyboard 8 1/0 slots Practical Computing, July 1978) $:E!uters (reviewed in computer 8KB ROM Attache Attache Chassis 8080 16KB RAM Video screen Built-in Parallel printer system based extra keyboard interface is on the S l O O extra.10 slot bus S l O O bus Bytronix Megamicro 808OA/ZSO 64KB RAM 12411 20 X 80 2 parallel ports Dual 8-in 120 cps black and white and 2 serial floppy disks printer Micro-computers Commodore KIM1 (reviewed 6502 1KB RAM 6-digit LED Small calcu- Built-in inter-Systems in Practical 2KB ROM display lator type faces for audio Division CompUring keyboard cassette and Commodore PET (reviewed Home corn- 6502 Full alpha- IEEE-488 Systems in Practical puter numericsand interface and a Division ComMina.graphics parallel user VDU ports (1MB) October 1978) teletype Octdber 1378 2001-8 8032 ke are. % L a t o r tport tape 32KB RAM Typewriter 32KB RAM $ ~ l ~ ~ o v D u Extra 1 32KB RAM Two RS232 Dual 8-in Compelec Compelec 280 Electronics Series Compucolor Compucolour I1 Packaged 8086 8KB RAM 1341-1 32 x 64 72 key RS232 port Built-in single serial ports floppy disks 1 parallel port (512KB) (reviewed in system eight-dour detachable mini floppy Piact+ June 1979) display with keyboard disk drive alpha-numerics (SIKB) + graphics ComWw, Compucorp 625 In the600 280 60KB RAM Built-in 9-in Full key- RS232 serial Dual Sf-in 40 cps series of desk- 16 x 64 boardwith I/Oport floppy disk printer top computers black and 40 user (700KB) white VDU definable keys Computer Mini Kit Kit 280 16KB RAM Serial and Minifloppy Centre arallel I/O disk drive E 100 mother board Computer System I Modular 32KB RAM 9-in 16 x 64 Dual 5t-in Workshop System black and floppy disks white VDU (170KB) Cromemco 2-2 280 16KB RAM Cromemco 2-2H ZSOA 64KB RAM RS232 serial llMB hard printer inter- disk drive, face 2 floppy disk drives Cromemco System I1 Factory Z80 32KB RAM Dual printer Dual mini-assembled interface floppy disks dguration of System 11) (System 11/64) (New con- (64KB) serial inieriace Cromemco System I11 280 32KB RAM Dual printer Dual 265 KB interface floppy disks RS232 &a1 face) (System III/64) (New con- (BOA) (64KB) (console and (dual 8-in disk figuration) printer inter- drives) Equinox Equinox 300 280 32KB RAM SlOO bus Dual rnini-Computer 48KB RAM floppy disks Systems 48KB RAM Dual 8-in floppy disk YSTEMS CURRENTLY AVAILABLE 261 given at the end of the table.Approximate price of basic system Maximum memory (excluding VAT) expansion Accessories available Software support and comments OD system Software it €65 assembled ‘9 8KB EPROM on a Up to 8KB RAM and fully buffered memory card Monitor and machine code programming. BASIC and development system for industry, DOS may be contained on the memory card electronic engineers Aimed as a low-cost 6503 as a control system for ;95 48KB RAM Single minifloppy disk (115KB) BASIC operating system, printers colour graphics PASCAL rom L1737 (full 64KB RAM Two single or double density Disk BASIC available isiness system 8-in floppy disk.VDU screen )prox. L5000) and parallel printer are extra to basic system rom €6080 BASIC FORTRAN COBOL, PASCAL all run under CP/M Aimed at business and university users :om €100 64KB RAM No software available 8KB PETS may be expanded to 32KB minifloppy disk-user storage system ASSEMBLER groups and software houses c.g., with expansion board 950KB; dot matrix PETSOFT Tape cassette drive dual 8KB BASIC in ROM operating Plenty of software from user is5 &SO 150 596 396 3 400 64KB RAM Double density floppy disk CP/M BASIC COBOL PASCAL, (lMB) 1KB EPROM FORTRAN ASSEMBLER rom L1390 32KB RAM Extended BASIC in ROM, ASSEMBLER System now ranks 4th behind PET TRS-80 and Apple in “personal” computer sales 60KB RAM ASSEMBLER BASIC Also a 655 model with 320KB floppy disks and 12-in, 20 x 80VDU 86 ALGOL Microsoft Also maxi-kit similar to mini-kit BASIC COBOL FORTRAN but with larger 8-in floppy disks 600 Printers ASSEMBLER BASIC FORTRAN From an extens ve range of FLEX PASCAL PILOT fully compatibl modular systems om €372 (kit) to 512KB 3 minifloppies or 4 8-in BASIC FORTRAN COBOL Aimed at OEM educational and OW+ floppies ASSEMBLERS industrial/scientific users 998 Software available dm €1995 512KB including two Up to 7 terminals FORTRAN IV COBOL Multi-user system for software additional floppies Extended BASIC ASSEMBLERS development word processor am €2995 to ooo+ multichannel A/D and D/A 2 additional disks 12KB, interface PROM programmer] up to 7 terminals Software as for System I1 L50 :95 CP/M BASIC.also available are F~RTRAN COBOL PASCAL maCrO-ASSEMBLER text sharing k40000+ processor With expansion it is a multiuser system for 2-12 users with tim 262 TABLE I-Hardware description of basic system Micro- Minimum Interfaces and Type of backing Manufacturer Unit Configuration processor* memory Visual display Keyboard 1/0 ports memory Other Exidy Sorcerer 280 Either 16KB 12-in video 79 key key- Cartridge and (reviewed in or 32KB monitor board 256 cassette inter-Practical RAM characterset face Computing (128 graphics) March 1974) Heath Data WH-89 All-inane Two Z8Os 16KB RAM “Intelligent” Floppy disk System computer video terminal Heath Kit H8 Kit SOIOA 4KB RAM “Intelligent” Octal key- Benton Harbor front panel board 60-pin bus VDU extra Built-in speaker + power supply ~~ Heath Kit H11 Fully wired 16-bit LSI 12KB RAM andtested CPU KDllF board with 16-bit LSI CPU Built-in paper tape reader ~~ Hewart Mini 6800 Micro- Mk I1 electronics 1KB user Upper and RAM.lower case 1KB monitor VDU with 1KB VDU graphics option RAM CUTS inter-face Hewart 6800s 16KB RAM Upper and Printer and Micro- lower case, electrouics graphics Hewlett- HP85 Compact 16KB RAM 54n 16 x 32 RS232 port Cassette 64 cps Packard portable black and 4 parallel (200K) printer Imsai VDP 40 32KB RAM 9-in display Standard (Serial I/O) Two5)-infloppy L$ZS2a-unit white VDU p o a keyboard disk drives.VDP 42 32KB RAM 9-in display Standard (Serial I/O) Larger disk screen screen keyboard capacity ~~ ~~ Intertec Super Brain Everything TWO ZSOAs 64KB RAM Full upper and Numeric and SlOO bus TWO double contained in 1KB 2708 lower case control key- density 5-in one unit PROM ASCII. 128 board floppy disks set with separate (160KB each, limited graphics. 320KB total) 84 user definable keys, program marker cursor and reverse video. 25 rows x 80 characters, 8 x Sin ITT ITT 2020 Single unit 6502 16KB RAM Extra Extra Luxor ABCSO Personal ZSOA 16KB RAM 12-in black and Professional 46/80 bus to Cassette Real-time computer 1KB picture white TV type key- connect peri- memory clock loud 2KB moni- screen displays board pherals Le. speaker tor ROM upper and floppy disk 16KB Basic lower case units printers in ROM digits.Built-in plotters, graphic mode measuring giving 64 instruments, different etc. V24 jack graphic symbols. for connecting Size 24 rows to a telephone of 40 characters modem to communicate with other computer systems via the public tele-phone network. IEEE-488 inter-face and an RS232 port Micronics Micros Kit or Z80 2KB RAM 47 key solid Interfaces for assembled state ASCII video (48 x 16 keyboard character video matrix), cassettes, printer and UHF TV. (RS232 serial I/O and 2 dual parallel 1/0 ports ontinued 263 Approximate price of basic system Maximum memory (excluding VAT) expansion Accessories available Software support and camments Software on system rom l760 without DU Micro lis floppy disks.BASIC ASSEMBLER EDITOR High-resolution graphics EPR& pack for your own word processor capability programs on cartridge Expandable i23 (including VAT) 65KB Disc drive Benton Harbor BASIC, extended BASIC text EDITOR applications ASSEMBLER console debugger Aimed at education and 1183 32KB PDP-11 software including BASIC and FOCAL EDITOR, relocatable ASSEMBLER linker, absolute loader rom €127 BASIC in ROM Aimed at education and small business .-om €276 Room for 8KB BASIC in ROM Software available 2 240 32KB BASIC Full dot matrix graphics 4 507 64KB Also available VDP 44 with 4 950 floppy disks (780K) 1996 Printer SlOO bus allows expansion with 2 DSDD 8-in floppy disk-2.4MB 8-120MB hard disk colour VDU plotter, modem main-frame interface, etc.CP/M disk operating system Disk and processor use separate 280 microprocessor hence computing can continue during disk operations 67 48KB single minifloppy disk (116KB). MONITOR ASSEMBLER extended RS232 port BASIC (ROM) 195 40KB printer plotter BASIC with resident EDITOR Aimed at small business and ASSEMBLER education markets it €360 assembled 64KB 2KB BASIC Aimed at small businesses and 199. System with process controllers :oustic coupler and DU Ll02 264 TABLE I-Hardware description of basic system Micro- Minimum Interfacea and Type of backing Manufacturer Unit Configuration processor. memory Visual display Keyboard 1/0 ports memory Other MicroV Microstar Singlebox 64KB RAM Three RS232 Twin 8-in serial inputs floppy disks Midwest MSI 6800 Kit or 16 KB ACT 1 Cassette inter-Scientific assembled terminal face Instruments Nascom Kit Z80 4KB RAM Full alpha- Parallel I/O, Micro- EzZZsin computers numeric serial data keyboard interface Computing, January 1978) National Pegasus Z80 48KB RAM 12-in CRT 58 key SlOO bus two Double density Multiplex keyboard serial inter- floppies faces one (320KB) parallel inter-face ~~ ~~ Netronics Elf I1 Single board 1802 256KB RAM (Video output Hex key- (Cassette interface + ard as an an option) computer option) option) Netrdcs Explorer 85 8085A 4KB RAM VDU board ASCII key- SlOO expansion board cassette.RS232, TT7 interface, on board I/O ports ~~ Newbear 7768 4KB RAM cassette + VDU North Star Horizon Integrated Z80A 16KB RAM Does not include a terminal instead uses I/O Single mini-interface disk hardware + computer printer drive (180KB) in one chassis port to a conventional VDU similarly for a floppy disk Ohio Challenger Supplied as Scientific C2-4P two boards with open slots for expansion Ohio Challenger Similar to Scient&c C2-8P C2-4P Ohio Challenger Triple pro- 6502A 32KB RAM Dual 8-in Scientific C3 cessor archi- Z80A floppy disks tecture 6800 ~~ Ohio Superboard I1 Everything 6502 4KB RAM 32 x 32 video Full 53 key Kansas City Scientific on a single 2KB monitor ipterface allow- keyboard standard audio board with in ROM 8KB ing upper and with upper cassette inter-a built-in Basic in lower case + and lower face for use with keyboard (not a kit) ROM graphics case an ordinary cassette recorder ~ ~~ Pertec System 1300 32KB RAM Serial interface Powerhouse Powerhouse 2 280 16KB RAM 5-in built-in Full key- RS232 interface Micro- VDU board comuu ten ~~ ~ Processor So120/16 8080 16KB RAM Video Integral S100 cassette Technology interface keyboard interface serial (including with numeric and parallel graphics) pad interface carp.Rair Black Box Basic system 2 program- Dual mini- Real-time in one unit mable serial floppy disks clock I/O interfaces 80KB each (can expand to 8 serial inter-faces) RCA Cosmac Kit or 1802 Hex keypad Output to TV Research 380-2 Z80A 16KBRAM RS232 port Cassette Machines assembled scree mtinued Approximate price of basic system Maximum memory Software support a d comments (excluding VAT) expansion Accessories available Software on system 950 (includes a STARDOS operating system Word processing software porter generator enables system to have 3 available w w e ) VDUs plus a fourth job running simultaneously s i c &1100 3 disk systems 32KB Minifloppy system with triple 516 as kit) memory with mini-floppy 56KB memory with large floppy or hard disk system lOMB driver of 8 bytes each large floppy system with 4 312KB disks hard disk system with 65 64KB Up to 64 parallel I/O ports Available as Nascom 2 with 8KB RAM and 8KB BASIC in ROM 700 (includes a -directional inter) 8-in drives 1-2MB additional drives digital recorder isic kit 80 64KB ;emblek&l b0, 3 board L35 (Machine code or Tiny BASIC option) om L297 (includes 64KB ogramxnable timer) Disk software microsoft BASIC on cassette used 8080 and 280 software can be .om k45 64KB (only in kit range of BASICS 95-L2 500 56KB 4 minifloppy disk drives Extended BASIC DOS form) (180KB each).12 slot S-100 bus enables one to plug in many types of peripheral boards including a hardware floating point board 4w1204 i5O-13 000 Multiple terminals) printers Can run virtually all 6502 Software available including 6800 8080 and 280 code 8KB microsoft BASIC in ROM word processing + data base management >m La98 Expander board with 24KB static RAM including an additional minifloppy interface port adaptor for a printer and 081488 line expansion interface 100-5 500 64KR 4 serial ports ~~ .SO (including 32KB I-time clock) to 30 14KB Basic X-Y graphics, 2KB monitor 50 excluding 6KB Complete floppy disk system Extended BASIC FORTRAN, :itor and cassette with word processing (&5000) FOCAL ASSEMBLER EDITOR m €2300 64KB ~~~ ~~ IMB disk storage (or lO.MB hard disk) range of peri-pherals Extended BASIC interpreter, FORTRAN IV and COBOL D O assembled ASSEMELER and machine code programming Tiny BASIC extra I8 56KB Dual 54-in or dual 8-in mini- Tiny BASIC extended BASIC High-resolution colour graphics floppy disks (168KB and lMB, respectively) CP/M COBOL APL FORTRAN ALGOL board availabl TABLE I Hardware description of basic system Micro- Minimum Interfaces and Type of backing Manufacturer UDit Configuration processor.memory Visual display Keyboard I/O ports memory Other ~ ~~ Rockwell Aim-65 Single board (reviewed in unit Pract+ ComMWA July 1979) 1KB RAM 20 characters Full alpha-16 segment numeric on-display board keyboard On-board thermal printer ~~ Use audio Science of Sinclair ZXSO Single unit 780-1 1KB RAM Use domestic Built-in Cassette Cambridge (reviewed in (available as TV touch- interface cassette Personal kit) personal sensitive television recorder computer computer World April 1980) keyboard. interface Graphic facilities SDS SDSlOO Single unit 32KB RAM 12-in video Keyboard Serial and display parallel inter-facing Semel Semel 1 Z80 16KB RAM 12-h' 24 x 80 Keyboard RS232 port Single floppy black and disk unit white VDU (250KB) 4KB monitor board and interface RS232 joystick, Sord MlOO ZSO 16KBRAM VDU Full key- Dual cassette Two-cham function pad port SlOO bus 24-h clock Sord M222 Z80 64KBRAM VDU Full key- Real-time 70KB mini-graphics board-md clock audio floppydisk numenc casseiteinter- drive keypad face two serii4 arts, three &o slots, interface for two external 6602 Memory available in Synertek Syml Fully assembled ~ and tested 4KB bl& up to 64KB.4KB ROM monitor Keypad Highspeed cassette inter-face (2400 baud) Tandy TRSSO Level 1 280 4KB RAM 12-h 16 X 64, black and white VDU Cassette ~~~ Tandy TRSSO 280 16KB RAM 1 2 - i ~ ~ 16 X 64 RS232 Cassette Level 2 black asd interface 1 white VDU parallel port vector MZ Z80 56KB RAM 3 serial ports Dual mini- PROM Graphics 2 parallel floppy disk burner ports (6SOKB) Zilog MCZ 1/05 zso 64KB RAM RS232 port Dual 8-in Debug in ~ ~ O D D V disk 3KB PR( (666KB) The following are the manufacturers of the micropr~~essors given in the table: Intel .. 8080 8080A 8086A 8086 MOS TechhOl~gy ' . . . MCSbOP M"cs6602k Motorola MC6800 NEC 780-1 RCAcoSm'a'c 1 : 1803 Zilog Z80 Z8OA 26 ontinued Approximate price of basic system Maximum memory Software support and comments (excluding VAT) expansion Accessories available Software on system <B €250 4KB ASSEMBLER EDITOR BASIC Compatible with Commodore SB €315 KIM ~~ 00 kit f80 16KB LOO for full expansion ~ ~ ~~ ~~~ ~ Requires a mains adapter 4KB BASIC in ROM EDITOR, operating system No facilities for trignometric, logarithmic or floating point arithmetic functions at time of press.Possibly included at a future date :om €3 750 46KB up to 8KB Line printer SD monitor program PROM Tom g2 900 64KB Printer light pen maximum of 8 x 250KB floppy disk units BASIC COBOL FORTRAN 126 48KB rom k3450 to €4123 Bar code reader. TMS-1000 Also M223 modelerne hardware ncludes desk and rinter) external Minifloppy 'disk drives 350KB minifloppy disk development system two as M222 plus single or double mm 0 6 0 64KB Port expansion kit TV inter- Compatible with KIM1 i.c. can face card teletype interface use same software 85 16KB BASIC ASSEMBLER BASIC in 4KR ROM. Level 1 is upgradable to level 2 60 48KB hfhiflOppy &Sk Unit numeric BASIC IIIaCrO-ASSEMBLER, pad line printer FORTRAN 595 DOS BASIC ASSEMBLER System B has graphics and ALGOL C BASIC COBOL, FORTRAN PASCAL numeric pad 200 ~ R10 o/s maCrO-ASSEMBLER, PASCAL BASIC COBOL FORTRAN, 26 268 BETTERIDGE AND GOAD THE IMPACT OF TABLE 11 SUMMARY OF ANALYTICAL APPLICATIONS OF MICROPROCESSORS Analyst Vol.106 Reference Application 1 2 3 4 6 6 7 8 9 10 11 12 Description of the hardware of a microprocessor and an application in electrochemistry Microcomputer for data handling in analytical chemistry Data processing in an analytical laboratory First microcomputer-controlled infra-red spectrometers (Perkin-Elmer, Models 281 and 283) Perkin-Elmer Model 460 atomic-absorption spectrometer with micro-computer Differential corrected spectra unit for the Perkin-Elmer Model MPF-MA fluorescence spectrophotometer Microprocessor-controlled automatic titrator Microcomputer-controlled single-task potentiometric titration system Real-time monitoring and control of gas chromatographs Microprocessor for data handling in gas or liquid chromatography Infrared spectrometer (Wilks Scien-tific Corp.Model 80 gas analyser) General-purpose microcomputer in analytical chemistry Description of application Two papers. The first deals briefly with the hard-ware principles involved in a general-purpose microprocessor. The second describes the use of a microprocessor in the construction of differential stripping electrochemical apparatus 1.On-line filtering of an analogue signal a t the output of a liquid chromatograph. 2. Use as a buffer memory to store rapidly data from a gas chromatograph - mass spectrometer and then output i t slowly to an inexpensive recorder. 3. Analyse noise from a new liquid chromatograph detector Microprocessor performs the data acquisition and reduction whilst the final evaluation and co-ordination are left to a larger processor Microcomputer directly controls the recording and optical systems i.e. gratings and filters are changed on computer command and both the mono-chromator and chart positions are tracked precisely Integration times of 0.2-60 s and scale expansions of 0.01-100-fold can be selected. It contains an automatic calibration graph correction feature Corrected differential excitation emission or excita-tion - emission spectra can be recorded The Intel 8008 microprocessor controls the titration and determines the equivalence volumes by locating the maximum in the differential ApH/AV versus V Successive determinations of calcium and magnesium in drinking water by complexometric potentiometric digital titration to two equivalence points.The functions of the microcomputer are : (a) to check the stability of the electrode signal and ultimately accept i t ; (b) to calculate and control the titrant delivery volume; (c) to establish whether inflection points occur and determine the exact location of those detected Intel 8008 and 8080 microprocessor-based systems were developed to control sampling operation of the gas chromatograph and perform real-time on-line data acquisition and analysis Microprocessor-based system consisting of an 8KB memory a display panel a cassette tape memory and an ADC is used for data processing.Tangential skimming method and the perpendicular method are used to process unresolved peaks in a tail of a main peak. Concentration of each component is calcu-lated automatically Microcomputer-controlled single-beam spectrometer can automatically analyse up to 11 compounds. Used to analyse quantitatively atmospheric pollu-tants Based on the Intel 8008 microprocessor its applica-tions include : 1. Use as a buffer memory to store rapidly data from a mass spectrometer and then output it slowly to an inexpensive recorder.2. Analysis of noise using a conventional recorder as an oscilloscope. 3. Normalisation and linear output of mass spectro-metric data grap March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 269 Reference 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Application Perkin-Elmer Sigma Series inter-active gas chromatograph instru-mentation Microprocessor applications in the process industry Single-cell protein production via a fermentor and a microprocessor Hierarchical approach to computers in the handling of time-critical problems in process control of experi-ments Multi-dimensional gas chromato-graph (Packard-Becker Model 429) Microprocessor-controlled photodiode array measurement system for analytical spectrometers Microprocessor-assisted calibration for a remote working level monitor Microcomputer for controlling chemi-cal analysis instruments Microprocessor - Ionalyser Automatic degassing of liquid samples for NMR and ESR spectro-scopy Microprocessor-controlled scanning dye laser for spectrometric analytical systems Application of microcomputers in chemical production plants Infrared spectrometer equipped with a microcomputer Portable neutron spectrometer/ Kerma rate meter Microprocessor-based multi-wave-length detector system for liquid chromatography Radiation monitoring system for use in a nuclear power station Description of application The series consists of a GC data system (mini-computer based) and four gas chromatographs that incorporate microprocessors.The microprocessors provide data processing control and monitoring, and fault diagnosis facilities Brief account of two applications involving micro-processors which were carried out at Warren Spring Laboratory Stevenage Hertfordshire Fermentor coupled to a microprocessor to control single-cell protein production. It regulates pH, temperature dissolved 0 and can take into account the 0 and CO in the exhaust gas. The micro-processor also logs and processes data Microcomputer shared between different experiments and plotting tasks under the control of a multi-programmed RC4000 computer resulting in a cut in resource demands in an on-line plotting system Apparatus consists of a basic gas chromatograph, modular flow system an analysis module and a controller unit based on an %bit microprocessor.The multi-dimensional approach allows analysis of systems that cannot be separated on a single column The Intel 8080A microprocessor with its associated hardware controls the functions of the 256 linearly spaced photodiode array system. It controls the scanning operation of the photodiode array selects a segment of the ultraviolet visible or near-infrared spectrum that is to be the active area of the photo-diodes collects and processes the data correcting for fixed pattern noise and background The microprocessor calculates 8-efficiencies and solves systems of linear equations with several unknowns. A review with no references A description of the Orion Model 901 Ionalyser and the ion-selective electrodes available for use with it Use of a microcomputer and apparatus for the auto-matic degassing of liquid samples by the cool and thaw technique before analysis by ESR or NMR spectroscopy are described The microprocessor controls the wavelength selection (from 360 to 650nm) by adjusting the angle of a diffraction grating and moves one of several dyes into the laser cavity.It can also collect data directly or be incorporated into a larger instru-mental system and run by a minicomputer Microcomputer use in production plants is discussed on the basis of actual and planned applications Wilks Model 180 infrared spectrometer is used for calculating the amount of various components present in a multi-component mixture An 8-bit microprocessor system was developed for use in a portable instrument capable of measuring and displaying energy spectra and Kerma rate.On-site analysis of data is possible and thus eliminates the need for a large-scale computer The microprocessor performs control data acquisition and data manipulation in order to optimise separa-tion. It digitises the analogue output of a linear silicon photodiode array A digital radiation monitoring system employing distributed microprocessors is describe 270 Reference 29 30 31 32 33 34 35 36 37 38 39 40 BETTERIDGE AND GOAD THE IMPACT OF TABLE II-continued Analyst Vol. 106 Application Instrument - computer interface based on the SDK-80 microcomputer Microprocessor in a pulsed NMR spectrometer Potentiometric stripping analysis Microcomputer-controlled automatic titrator Applications of microprocessors to analytical spectrometry Microprocessor-assisted high-precision viscometry Microprocessor-controlled differential titrator Microprocessor-controlled potentio-stat for electrochemical measure-ments Microprocessor to solve pH control problems Microcomputer-assisted single beam, photoacoustic spectrometer system for the study of solids A microprocessor on-line to general analytical instruments Microprocessor controlled HPLC with CRT display Description of application NMR spectrometers were interfaced through a Fabritex 1074 computer to a DEC System-10 computer by an SDK-80 microcomputer.The system has minimum hardware design is easy to operate and has LED program state indicators A low-cost pulsed NMR system was built designed around the Intel 8080 microprocessor.The micro-processor supervises instrument control and data acquisition. Data are stored in RAM and can be transferred over a telephone line to a remote time-shared computer for further processing Microcomputer controls the equipment for auto-mation of potentiometric stripping analysis. The higher resolution of time compared with manual equipment increases the sensitivity of the method An Intel 80/10 microcomputer programmed in BASIC controls an automatic titrator. Either fixed or variable titrant additions can be performed and a stable pH reading is ensured. The titrant can be added in volume increment as small as 5 x lo-' ml.The time taken to collect each point is recorded Three systems have been designed and integrated together into a hierarchical approach to demonstrate the diverse roles that microprocessors can play in analytical instrumentation : The systems are: 1. a data acquisition module to perform frequency ratio measurements ; 2. a monochromator controller; and 3. supervisory system programmed to link the first two units to it and form a hierarchical system for controlling a spectrometric system Microprocessor is coupled to a high-precision high-temperature torsion pendulum viscometer. Photo-detector signal level transition times are stored in the microcomputer and after some calculation transferred to a PDP-I1 minicomputer where the main viscosity computations take place Control of titration procedure and recording of data is performed by the microcomputer (Intel MCS-80) The system consists of a CPU 16KB RAM peripheral interfacing a timer an ADC and a DAC.It is intended for the control and unattended operation of a standard laboratory potentiostat Development of a microprocessor-based pH control system for batch processing is described Expensive commercial data acquisition units have been replaced by commonly available inexpensive and easily constructed components. The use of the microcomputer for data acquisition and reduction has eliminated the problems associated with the single-beam mode of operation i.e. source output correction background compensation etc.Hewlett-Packard 9825A microprocessor was used to collect and process data from smaller commonplace analytical instruments e.g. electronic balances, atomic-absorption spectrometers and chloride meters, which are generally characterised by their slow (operator-limited) data rate. The incorporation of a microprocessor keyboard and a CRT in the Varian Model 5020 liquid chromato-graph offers the user a simplified approach to instrument operation. The CRT permits continual display of instrument status which eliminates the need for status lights and indicator March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 271 Reference Application 41 42 43 44 45 46 47 48 49 50 61 62 Microcomputer-controlled potentio-metric titration system for equili-brium studies Microprocessors and the chemist Microprocessor-based liquid chroma-tograph with CRT display (Varian, Model 5000 high-performance liquid chromatograph Applications of microcomputers to analytical chemistry Microprocessor-based linear response time low-pass filter Microprocessor-based system for measuring low-frequency (100 Hz) electrical noise spectra Microprocessor Mossbauer spectro-meter Microprocessor-controlled ultra-violet - visible spectrometer with incorporated recorder (Perkin-Elmer, Model 564) Microprocessor-based process gas chromatograph Microcomputer-based gas chromato-graphy system A review of microprocessors in laboratory instrumentation General-purpose microcomputer in laboratory automation Description of application The microcomputer is based on an Intel 8080A microprocessor with the control programme being stored in PROM.The system delivers an optimised volume of base that corresponds to uniform changes in electrode potential throughout the entire titration A review on microprocessors with current and possible future applications for the chemist The use of the 3-solvent capabilities of the Model 5020 chromatograph for the separation of polystyrene oligomers and the advantage of an automatic loop valve of the Model 5000 chromatograph in moni-toring the course of a chemical reaction are discussed A review with 34 references The filter has a response time half that of an equivalent RC filter and is therefore particularly suitable for application in analytical instrumentation (?.g.atomic absorption) where reduced response time means reduced sample consumption shorter analysis time and reduced effects from drift The microprocessor determines the noise auto-correlation function by using simple one-bit auto-correlation arithmetic ; subsequent Fourier trans-formations to find the power spectra are performed on a larger computer. Current noise measurements in insulating polymers are described A Mossbauer spectrometer based on the Motorola MEK 6800D1 microprocessor kit is described; i t provides a simple cheap and flexible alternative to conventional multi-channel analyser systems The apparatus contains a background compensator to provide extremely straight base lines a built-in calculating system to control the automatic recorder, automatic standardisation process a keyboard to input appropriate parameters and an automatic cuvette charger.First and second derivatives of spectra may be used to solve difficult spectroscopic problems and with integration measured values can be determined accurately and recorded under unfavourable conditions A data processor unit controls from 1 to 6 chromato-graph analysers. Each analyser contains a micro-processor that controls all of the functions associated with the analysis performs all measurements and integrates the area or finds the peak height. Only final data for selected components are transmitted to the data processor The gas chromatography system has an interlocking microcomputer for regulating temperature timing integrator and recorder A small table consisting of five typical micro-computing systems with their respective price, memory interfaces etc.Also includes four refer-ences three on hardware aspects and one on microprocessors in instruments A general-purpose microcomputer data system was developed to bring on-line individual chemical instruments. The design features include modu-larity optimised 1/0 structure for real-time experi-mental control and data acquisition flexible video display generation for operator intervention and data presentation etc 272 BETTERIDGE AND GOAD THE IMPACT OF TABLE II-continued Analyst Vol. 106 Reference Application 53 54 55 56 57 58 59 60 61 62 63 64 Versatile microcomputer-controlled titrator Microprocessor applications in the analysis of natural gas and cement kiln gas Microprocessor-controlled electro-chemical system Automatic analysis of organic pollu-tants in water via a microcomputer-controlled GC - MS system Quality control of polymers using a microprocessor-based thermal ana-lyser Microprocessor system to monitor distillation plate efficiencies Thermal analysis of plastics via microprocessor control Microcomputer-controlled buffer gradient for ion-exchange chromato-graphy Nuclear criticality evacuation with telemonitoring and microprocessors Microcomputer control of systems in chemical engineering Uses of a desk-top computer with graphics in automated geochemical analysis The implementation and operation of microcomputers in analytical laboratories are discussed Description of application The titrator is capable of handling ultraviolet -visible spectrophotometric potentiometric and amperometric end-point detection and acid - base, redox complexometric and precipitation reactions.The ADD 8080 microcomputer system is used which is based on the Intel 8080A microprocessor chip. It controls titrant delivery pre-titration adjust-ments end-point measurement equivalence point determinations calculations and display of results in the derived units. All titration programs were written in BASIC and stored in paper-tape form Determination of the octane number the dew and frost points H,S the natural gas calorific value and combustion control Based on the Motorola M6800 it is designed as a software orientated system with a minimum of hardware parts.The system performs cyclic voltam-metry but can easily be adapted for controlled-potential electrolysis chronopotentiometry or other electrochemical techniques The system prepares the sample and processes the analysis data. The sensitivity is 60 p.p.t. of organic compounds A Du Pont R90 modular microprocessor-based programmer was used for quality control testing using differential scanning calorimetry thermograms An optical probe and microprocessor monitor were used for determining bubble size and velocity in a froth formed on a sieve plate and for evaluating plate efficiency on-line. The technique was used to study the behaviour of surface tension positive and negative binary systems and the effect of liquid viscosity in gas absorption Three techniques involving microprocessor control are measurement of total volatiles in compound resins ; separation of extenders or plasticiser carbon black and inert filler from polymer and elastomer formulations ; and separation of fillers This system was developed for switching between several buffer solutions to produce a mixture with independent elution gradients for two ions.It was used as a gradient programmer for the separation of amino acids by ion-exchange chromatography. The microcomputer also supervises the automatic facili-ties of an amino acid analyser The criticality system a t Argonne National Labora-tory is described. The site microprocessor maintains a current data table detects faults drives a printer and communicates with the central telemonitoring station A review with six references Tektronix 4050 series graphics microcomputer was used as an on-line data processor in X-ray fluores-cence spectrometry.During calibration plots of X-ray intensity versus elemental abundance in the calibration standard samples were available for on-screen display less than 1 s after depressing a user-designated key. Ten-element iterative matrix correction calculations require less than 3 s per sample A revie March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 273 Reference Application 65 66 67 68 69 70 71 72 73 74 75 76 Microprocessor-assisted quality con-trol of the external CPS 24 GeV/c beams A general purpose electroanalytical system based on a microcomputer-controlled coulostatic generator Microcomputers in laboratory auto-mation Use of a sequence controller in process control A microcomputer-compatible method of resolving rate constants in mixed first and second order kinetic rate laws Practical microcomputer technology A lightweight microprocessor-con-trolled solar neutron detector Microcomputer-controlled potentio-metric analysis system Microprocessor-based millivoltmeter in applications of ion-selective elec-trodes Microcomputers interfaced t o a high-resolution NMR spectrometer, equipped with a digital integrator Microcomputers in the wet-chemistry laboratory Microprocessor-controlled photon counter for pulsed optically detected magnetic resonance Description of application To minimise beam loss and to retain beam quality a fine control of the CERN p synchroton beam positions is required which is achieved by a micro-processor in the CAMAC system This consists of a multimicrocomputer system as the control logic to control the charge generation (which adds charge to the working electrode) and the voltage measurement system (which monitors the potential).The multi-microcomputer system con-sists of three independent microcomputers each responsible for a specific task. One co-ordinates the operation of the other two so that the system works as a functional unit. The electrochemical system was applied to the polarographic determina-tions of cadmium.The sensitivity of the system was limited to 0.5 ~ L M of cadmium. Block diagrams of microcomputers used as communi-cation controllers device controllers and terminal adapters for 37 pilot plants are given A review with no references Both the first- and second-order rate constants and t = 0 are calculated from absorbance versus time data. The method is except for one minimisation, a one-pass process and is therefore easily used on a microcomputer or programmable calculator A review with four references The microprocessor controls the data acquisition, accumulation of histograms and the encoding of data for the telemetry systems The use of CONVERS (a language based on an interpretative compiler) in microcomputer instru-mentation combines the high-level software capabili-ties of a minicomputer whilst retaining the ease of construction and cheapness of a microcomputer-based system.Application of this system to end-point determination in potentiometric titrations and fully automated generation of calibration graph data for multiple ion-selective electrodes are described A commercially available microprocessor-based millivoltmeter was used as a simple method for converting appropriate millivolt readings to sample concentrations Multi-scan averaging is used to improve signal to noise characteristics and automatic calculation of the percentage purity of the sample is performed. It was successfully used to determine organic compounds quantitatively A microcomputer with an analogue to digital con-vertor is interfaced with an analytical balance a device for amplifying potentials a photometer and a motor-driven burette.The microprocessor per-forms its tasks via eight programs controlled titration incremental equilibrium titration titration a t pre-selected end-point pH titration dosing, sample dissolution with definite concentration, dilution and conditioning The multi-accumulator photon counter is capable of sequential counting over as many as four intervals. The microprocessor which controls the counter, manipulates the photon counts during each of the duty cycles. The instrument is inexpensive yet versatile and is suited for other applications in which sequential counting is necessar 274 BETTERIDGE AND GOAD THE IMPACT OF TABLE II-continued Analyst Vol.106 Reference Application 77 78 79 80 81 82 83 84 85 86 Microcomputer on-line to a Moss-bauer diffractometer Device for demonstrating micro-processor interfacing Basic concepts and terminology of microprocessors New application possibilities for microprocessors in rapid kinetics in clinical chemistry Development of a microprocessor-based system for simultaneous data processing of different automatic elemental analysers Microprocessor-controlled liquid chromatograph - atomic-absorption sampling system Pulsed-gradient NMR diffusion measurements with a microcomputer Microcomputer-controlled titrations Practical aspects of microprocessor-based digital controller systems Microcomputer-aided high-speed potentiometric titration system Description of application This was devised to separate the elastic scattered intensity and the inelastic one in the vicinity of Bragg reflections.A selective modulator to move a transducer at a resonant velocity and a t a non-resonant one alternatively is used for automatic separation The design of a prototyper for use in interfacing between a microprocessor or minicomputer and an external device is described. Detailed diagrams and a component list are included A review with no references A review with references of the clinical use of kinetic methods in combustion with built-in microprocessors The microcomputer allows users already in possession of multimeters or integrators an electronic balance and elemental analysers to combine their equipment into an automated system.After the analysers have been started the system processes all data and calculates the blanks calibration factors and per-centages successively Description of a system for trace analysis of metal-containing compounds. It consists of a Motorola 6800-based Heath Kit microprocessor trainer an external clock an electronic interface to the dispenser and to the atomic-absorption spectro-photometer and the dispenser system A microcomputer was adapted for NMR diffusion measurements using the pulsed field-gradient method. This made possible the use of samples with faint spin echoes The regulation of the titration process and the detection of the end point by means of a micro-computer and the construction and software of a process autotitrator are described A review with no references Titrant additions and time intervals were chosen and controlled by the microcomputer which also evaluated the equivalence volume based upon linear titration plots by using four titration points before the equivalence point.The apparatus was capable of determining the concentration of strong and weak monoprotic acids in less than 22 s with relative standard deviations of 0.1-0.2% titrator by acquiring feedback to the microprocessor which would then control the rate of addition of titrant until a pre-set potential was reached and taken as the equivalence point; (ii) to use a linear plot of the titration data to determine the equivalence point (Gran's method); (iii) to use a differential method to locate the equivalence point.The first of these is straightforward and can lead to a practical improvement in existing instrumentation. However we rejected it on the grounds that there was little novelty in such an adaptation and that it did not make full use of the potential of the microprocessor. The use of Gran plots seemed to have many attractions.8' They have been advocated for several years as an accurate way of determining the equivalence point but the transformation of titration data into the functions plotted e.g. volume x vs. volume are tedious to perform by hand. The microprocessor can perform the necessary conversions on the run and calculate the least-squares fit for the linear plot that should result. Further most of the points used in the calculation are taken from the mid-part of the titration curve thus minimising the problems arising from fluctuations of signal that occur near the equivalence point in a potentiometric titration March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 275 In practice it was found that the curves deviated from linearity near the equivalence point and even refinements of Gran’s equations such as that due to Johanssonss [equation (l)] and Pehrsson et aZ.s9990 [equation (2)] were not much more successful.v e - v = K ( [H]V + v { H } [HI = F { H } [OH] + v { [ H ] - [OH]}) (1) where r = V v } + ‘ F { H } [HI = - vo ‘{H} [OH] C and V is the equivalence volume Vo is the original volume of acid V is the volume of base of concentration C added K is the acid dissociation constant and the brackets round species distinguish activity and concentration terms in the usual way.The key to the difficulty is seen in the need for an accurate value for the dissociation constant and in the need to sort out the activity and concentration terms. This can be done by further computation but it is going away from the simple relationships that are more suitable for the limited computational capability of a microprocessor. The third approach the use of a differential of the titration curve was then investigated. It had the disadvantages that the peak in the differential curve concides with the equiva-lence point only when the titration curve is symmetric and it makes most use of the data close to the equivalence point.These disadvantages were offset by the ease of computation and the applicability of the algorithm to many systems no knowledge of K being required. I t was found that the deviations from the equivalence point with asymmetric titration curves was not significant (or could be corrected empirically with a titration factor) and that a simple smoothing routine could reduce the effects of erratic signals close to the equivalence point. (A more effective solution to noise was to reverse the direction of stirring, so that the titrant was better mixed before it reached the electrodes.) All of these calculations were performed via a terminal on a mainframe computer with a program written in a high-level language such as BASIC. Consequently the time involved in the development of the best solution to the problem was used more economically than if it had been written directly in machine language.The program had (in 1976) to be converted into assembly code,91992 a good reason for using the simplest adequate program. As it was to be used for a specific application the algorithm in essence specified the addition of a fixed volume of titrant. During the addition which was to be made in fixed increments a t a fixed rate (controlled and counted by the microprocessor) the titration data were collected and smoothed and the difference between successive points was calculated and stored. At the end of the run the two largest differences were found and associated with the two equivalence points of the sequential titration. A more exact value was calculated by inter-polation of the five points around each peak.The values were checked to ensure that they fell within reasonable bounds and were then displayed. The full titration assembly is shown schematically in Fig. 1. The automatic burette pH meter and electrode assembly were taken from a conventional autotitrator. The control functions were exerted as a pulse to a relay and the analogue to digital conversions and vice versa were carried out by cheap solid-state devices. Programming the microprocessor in assembly code took some time and was facilitated by the availability of a development system that included fast punched-tape readers program debugging facilities and PROM simulators. Typically a complete microcomputer development system MDS costs approximately QOOOO.The microprocessor was an Intel 8008 276 Microprocessor r BETTERIDGE AND GOAD THE IMPACT OF DAC - Osci I loscope L Analyst VoZ. 106 Y I I I In t - n tzl output Fig. 1. Arrangements of apparatus for titrations controlled by the microcomputer. As a laboratory instrument it proved very satisfactory. Two titrations could be per-formed in sequence in 1 min with a coefficient of variation of 1%. If 4 min were taken the precision improved to 0.25%. At a higher rate there was a slight systematic error but a t the slower rate the accuracy was as good as with a manual titration. The elimination of mannitol from the procedure resulted in sufficient savings in the course of 1 year to pay for the material cost of the apparatus. The approach of Christiansen et a1.8 was similar except that they used the titration data to slow the rate of addition of titrant as the equivalence point is approached.If one were to carry out a similar exercise now the major change in procedure would be to do the program development work on a PET or similar microcomputer with a suitable interface. This would automatically carry out the tedious step of converting an acceptable high-level computer language (BASIC) to machine code and effectively eliminate the develop-ment system that is needed to support the microprocessor. However even a PET linked to the basic components of an automatic titrator represents a laboratory “lash-up.” Control Laboratory Model required a further definition of objectives. The conversion of the laboratory prototype to a robust control laboratory instrument (i) There should be as few control knobs or switches as possible; ideally the operator should only have to press a run button and read off the result; (ii) it should be as compact and robust as possible by taking full advantage of solid-state electronics ; (iii) it should be applicable to more than one titration; (iv) it should have error-detecting routines that indicate common malfunctions such as broken connections ‘and defective electrodes.At this point additional advantages of the differential routine for the location of the equivalence point became apparent it was applicable to all non-linear titrations and it enabled one to exchange the expensive calibrated pH meter for a voltmeter chip costing about k25 and occupying very little space.The original program had to be modified to detect the equivalence point straight away instead of assuming a fixed increment of titrant. This was done by a moving window in which the signs of the difference between successive points on the differential titration curve were noted. As the equivalence point is approached, the window shows every difference as positive (i.e. an increment in ApH/AV) but as the equivalence point is passed the differences become negative (i.e. a decrease in ApH/AV). Some care is needed to discriminate noise spikes from a sharp end-point and to ensure that a poor end-point is detected. Economy suggests the use of only one readout and this is used to display the equivalence point the voltage if the We held the following to be important: Additional programs weIe needed to provide the error checks March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 277 appropriate button is pressed and an error-identifying number if the apparatus is mal-functioning.An additional control loop is used to re-fill the burette and switch on a “ready” light when the titrator is primed to re-start. When tested the program is placed in ROM where it remains fixed unalterable and immune from erasure by power failure or other electronic crises. These alterations to the laboratory prototype and the program listing took almost as long as the initial development. Other Approaches The strategy outlined above in both stages of development places emphasis on simplicity and implies a willingness to dedicate a microprocessor to a single function.We believe that this emphasises the quality of a microcomputer that distinguishes it from a minicomputer, i e . its smallness. At the outset of the work there was a severe limit on the available memory of a microprocessor (1-2KB) and this in a practical way endorsed the value of a simple approach. Large-scale integrated circuits (LSI) have become so available that with a PET or similar microcomputer one can obtain many of the facilities of a minicomputer and have by the standards of Intel 8008 immense computing power. Even with more advanced microcomputer kits such as the SDK 80 more complicated programs can be used and interaction between the operator and the computer is possible. One such system has been described by Leggett41 for the determination of equilibrium constants and others by Avdeef and B ~ c h e r ~ ~ and Martin and Frei~er.’~ The latter authors’ program allows the user to check on and correct for electrode drift.However the ultimate in interactive systems must surely be that of Wu and Malm~tadt,~~ the block diagram for which is shown in Fig. 2. However this is no longer true. DATA ACQU I S IT1 0 N Encoder pulses i Constant + current generator \ 1” Electrode L h amplifier Current (charge) Titrant delivery Fig. 2. Block diagram of the Wu and Malmstadt system.53 This system permits the operator to select virtually any type of titration and any appro-priate program for processing the titration data. Both linear and logarithmic titrations are permitted.However although the choice of procedure is left completely to the user the implementation is computer controlled and all of the results reported are satisfactory for most purposes. The flexibility is gained at the expense of simplicity and cost 278 BETTERIDGE AND GOAD THE IMPACT OF Analyst Vol. 106 In conclusion the only thing certain about microprocessor-controlled titrations is that most workers prefer a differential procedure for the determination of the end-point. There are completely different viewpoints as to whether it is preferable to have fixed or interactive programs. To some extent this depends on whether one is opting for a versatile laboratory instrument or one for process control. By now most of the possibilities have been tried and the consumer can feel confident that any commercial microprocessor-based titrator is reliable.In view of the development cost if he were to make it himself he may also feel it is good value for money. As in all computer applications the really difficult part is getting the chemistry right. Conclusion In such a rapidly growing area it is difficult to predict the future but a report of the University of Sussex Science Policy Research Units3 is of interest in that it takes a broad view of the future development of microelectronics. It predicted that technological advances will bring the cost of a microprocessor down to $1 by 1980 and 10-20 cents by 1985. It notes that “the reductions in cost have been made possible by the increased complexity of integrated circuits” ; however “the technology is now reaching a level where the majority of current and forseeably high-volume applications could be satisfied by single-component systems so that the demand for further increases in complexity is expected to decline leading to a stabilisation of technology.The future pattern of microelectronics is expected to show an increasing divergence between very complex circuits offering high performance at high cost and high-volume circuits where the emphasis will be on minimum cost.” The laboratory computer market is very small compared with the automobile industry, which is a major consumer of the microprocessors currently produced. However it seems reasonable to assume that easily programmable microcomputers for control computation and report writing will become a standard feature in most analytical laboratories.The pace of these developments has resulted in such a proliferation of books on micro-processors that it is impossible to list them all. Consequently we limit our recommendations to a very good general introduction to the fields4 and a recent book written specifically for the chemist by one who has been concerned with using microprocessors in commercial chemical instrumentation especially mass spectrometer^.^^ Assembly code Backing store Bus Cassette . . Central processor Central processing unit (CPU) . . Chip . . Glossary Instructions written in a machine orientated language. A device for storing programs or data externally from the main memory of the computer.Usually magnetic disks or tapes; access to the stored programs or data may take seconds as in the case of disk or minutes, as with tape. 1. A circuit over which data or power is transmitted. 2. A path over which information is transferred from any of several sources to any of several destina-tions. A magnetic tape in a plastic case that can be loaded and unloaded without manual threading of tape. Like a music cassette tape but holding programs or data which are accessed sequentially. The part of a computer system that performs compu-tations oversees the use of main memory and monitors input and output operations. See Central processor. See Silicon chip March 1981 Daisy wheel Debugging Disk . a Floppy disk Hard copy Hierarchy .. MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 279 . . Circular typing element with spokes around the circumference each with at least one raised character. Available in a range of type styles. Vying for popu-larity with the IBM “golf-ball” element. . . The process of isolating or correcting all malfunctions and/or mistakes in a piece of equipment or a program of operations. . . A recording medium originally developed for main-frame computers when it holds many millions of characters. A random access device in which the access time for any given location on the disk is constant . . . A flexible disk recording storage medium developed for minicomputors and often used with microcomputers. . . A machine printed document that can be handled and read by people without using special devices; usually program or data printed on paper.. . A specified rank or order of items. High-level language . . Image-sensing devices . . Inputloutput . . Instruction set . . Interactive display . . Interface . . Mainframe Main memory . . Microcomputer . . Microprocessor . . Minicomputer . . Processing unit . . . . A computer programming language that is less depen-dent on the limitations of a specific computer; languages common to most computer systems e g . ALGOL, FORTRAN COBOL. . . For example a detector that can collect 250 points in milliseconds. . . Commonly called I/O. A general term for equipment used to communicate with a computer. The process of transmitting information from an external source to the computer or from the computer to an external destination.The set of instructions that a computing or data-processing system is capable of performing. Any display that allows the user to input data in response to the information displayed. A common boundary between two systems (or devices) or parts of a single system. . . . . . . . . Large computer system (typically costing more than ~200000). . . Another name for the internal storage of a computer. . . A simple silicon chip containing a microprocessor as well as memory and some input/output communicating facilities. . . The central processing unit of a computer implemented on a single silicon chip. Originally a computer significantly smaller in size, capacity cost and software capability than the larger mainframe computers with which it is contrasted.The differences have been blurred by technical advances. . . . . See Central processor 280 PROM RAM ROM Shared facility Shared logic Silicon chip Simulate . . Software . . Stand alone Store . . Terminal . . Time sharing . . . . . . . . BETTERIDGE AND GOAD THE IMPACT OF Analyst Vol. 106 . . Visual display unit (VDU) . . . . . . Acronym for programmable read-only memory. It is permanent non-volatile memory; a program stored in PROM is not lost after power is removed and is ready for running as soon as the microcomputer is powered on thus eliminating the need for program loading.Although an acronym for random access memory ( L e , any location in this type of memory may be interro-gated without prior access to previous locations) this usually refers to read/write memory where memory locations may be read or altered by the user This memory requires power to retain its contents. Acronym for read-only memory. As the name suggests the contents of this memory may not be altered by the user nor is it dependent upon power being supplied to retain its contepts. I t has how-ever random access of memory locations. A series of semi-autonomous workstations sharing some common facilities normally a large disk store and/or printing capabilities. A configuration where all the workstations in a system are connected to a minicomputer that provides the processing power storage and printing facilities.A number of wafers of silicon fused together each containing hundreds or thousands of electronic circuits. To represent the functioning of one system by another, e.g. to represent one computer by another to represent a physical system by the execution of a computer program. The collection of programs i.e. sets of instructions, associated with a computer. A single independent workstation. See Memory. A device designed to send data to or receive data from a computer. Normally a typewriter-type key-board and VDU. A computing technique in which numerous terminal devices can utilise a central computer concurrently for input processing and output functions.Contains a screen similar to a television which displays text or data as it is entered on the keyboard or recalled from store. We are grateful to the SRC and AERE Harwell for the award of a Studentship to T.B.G. References 1. 2. 3. 4. 5. 6. 7. 8. 9. Dessy R. E. Janse-Van Vuuren P. and Titus J. A. Anal. Chem. 1974 46 917A 1055A. Goedert M. Wise S. A. and Juvet R. S. Jr. Chromatographia 1974 7 539. Weber O. Chromatographia 1974 7 547. Geary S. Instrum. News 1975 25 5. Slavin S. Instrum. News 1975 25 3. Anacreon R. E. Instrum. News 1976 26 6. Betteridge D. Dagless E. L. David P. Deans D. R. Penketh G. E. and Shawcross P. Analyst, Christiansen T. F. Busch J . E. and Krogh S. C. Anal. Chem. 1976 48 1051. Donaghey L. F. Bobba G. M. and Jacobs D.J. Chromatogr. Sci. 1976 14 274. 1976 101 409 March 1981 MICROPROCESSORS ON ANALYTICAL INSTRUMENTATION 28 1 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. Sato T. Takimoto S. and Kosaka I. Bunseki Kagaku 1976 25 94. Syrjala R. J. Environ. Anal. 1977 111. Wise S. A. Diss. Abstr. Int. B 1976 36 6117. Baudean J. E. Ettre L. S. Hartigan M. J. Hoberecht H. March E. W. Pigliucci R Purcell, J. E. and Welton B. Am.Lab. 1977 9 95. Fowles A. J. Chem. I%d. (London) 1977 6 214. Kryze J. Cordonnier M. Kernevez J. P. Lebeault J. M. and Pahaut P. Comput. AppZ. A n d . Chem. Data Plants 1977 252. Lindgard A. Sorensen P. and Oxenboll J. J. Phys. E 1977 10 264. Lips J . E. Am. Lab. 1977 9 63. Lovse D. W. Diss. Abstr. Int. B 1978 38 4771. McDowell W. P. Keefe D. J. Groer P. G. and Witer R. T. IEEE Trans. Nucl. Sci. 1977 24 639. Merlid K. Kjemi 1977 37 25. Nikkaki Co. Ltd. A + R 1977 15 562. Nishikawa T. Horiuchi Y. and Takeda I. Bunseki Kagaku 1977 26 731. Perry J . A. Bryant M. F. and Malmstadt H. V. Anal. Chem. 1977 49 1702. Pfeffer W. Regelungstech. Prax. 1977 19 258. Santo K. A + R 1977 15 499. Silberberg J . L. IEEE Trans. Nucl. Sci. 1977 24 386. Titus C. A. Diss. Abstr.Int. B 1977 38 1172. Warman E. A. Ingeneri S. M. Newman B. M. and Sowdon T. L. Nucl. React. Shielding 1977, Abramson M. J. and Goldstein J. H. Comput. Chem. 1978 2 139. Ader R. E. Lepley A. R. and Songco D. C. J. Magn. Reson. 1978 29 105. Anfalt T. and Strandberg M. Anal. Chim. Ada 1978 103 379. Avdeef A. and Bucher J. J. Anal. Chem. 1978 50 2137. Avery J. P. Diss. Abstr. Int. B 1978 39 2253. Borgen 0. S. Sandbu O. Anal. Chim. Acta 1978 103 389. Busch N. Freyer P. and Szameit H. Anal. Chem. 1978 50 2166. Cohen M. I. Heimann P. A, J . Res. Nut. Bur. Stand. 1978 83 429. Fraade D. J. Adv. Instrum. 1978 33 55. Harry E. and Stuart J. D. Anal. Chem. 1978 50 587. Jones M. E. Gidley G. C . Averill R. M. Proc. Anal. Div. Chem. Soc. 1978 15 47. Johnson E. L. and Luchetti S.J. Am. Lab. 1978 10 51. Leggett D. J. Anal. Chem. 1978 50 718. Littler J . S. and Reeves R. M. Chem. BY. 1978 14 118. Majors R. E. Varian Instrum. Appl. 1978 12 10. Nakayama T. Bunseki 1978 5 311. O’Haver T. C. Anal. Chem. 1978 50 676. Pender L. F. and Wintle H. J. Rev. Sci. Instrum. 1978 49 1968. Player M. A. and Woodhams F. W. D. J. Phys. E 1978 11 191. Spreitzhofer E. GIT Fuchz. Lab. 1978 22 117. Villabos R. Porter R. G. Leblanc R. and Hearn R. Proc. Annu. Symp. Instrum. Process Ind., Watanabe C. and Tomita H. Kagaku No Ryoiki Zokan 1978 120 143. Whitbread M. Chem. BY. 1978 14 560. Woodward W. S. and Reilley C . N. Pure Apfil. Chem. 1978 50 785. Wu A. H. B. and Malmstadt H. V. Anal. Chem. 1978 50 2090. Bailey S. J. Control Eng. 1979 26 61. Barringer G.E. Jr. Diss. Abstr. Int. B 1979 40 1145. Beggs D. Nut. Bur. Stand. (U.S.) Spec. Publ. 1979 519 169. Blaine R. L. Gill P. S. and Hassel R. L. SOC. Plast. Eng. Tech. Pap. 1979 25 822. Calderbank P. H. and Pereira J. Inst. Chem. Eng. Symp. Ser. 1979 56 2.3/27. Cassel B. Hall J. and Dander T. Soc. Plast. Eng. Tech. Pap. 1979 25 819. Davies A. M. C. Prescoh E. H. A. and Stansfield R. J. Chromatogr. 1979 171 117. Fergus R. W. and Moe H. J. Instrum. Aerosp. Ind. 197,9 25 305. Goto Y. Kagaku Kogaku 1979 43 518. Gunn B. M. Chem. N.Z. 1979 43 171. Haarmann T. and Koeser H. J. K. Fresenius Z . Anal. Chem. 1979 296 18. Heinze W. and Simon D. J. Tr Mezhdunar Konf. Uskorit. Zaryazhennykh Chastis. Vys. Energ., Hourdakis S. Diss. Abstr. Int. B 1979 40 707. Kaufmann W. E. Chem. Eng. Prog. 1979 75 49. Kawashima K. Kagaku Kogaku 1979 43 528. Kelter P. B. and Carr J. D. Anal. Chem. 1979 51 1828. Kitamoto A. Nishida N. Fujii K. and Inokuchi S. KugaPu Kogaku 1979 49 506. Koga R. Albats P. Frye G. M. Jr. Schindler S. M. Denehy B. V. Hopper V. D. andMace, Martin C. R. and Freiser H. Anal. Chem. 1979 51 803. Moody G. J and Thomas J. D. R. Lab. Pract. 1979 28 125. Morley F. O’Neill I. K. Pringuer M. A. and Stockwell P. B. Anal. Chem. 1979 51 579. Neumann G. Lab. Praxis 1979 32 35. Power R. K. and Nishimura A. M. Rev. Sci. Instrum. 1979 50 969. 434. 1978 33 39. loth 1977 2 356. 0. B. Nucl. Instrum. Methods 1979 165 47 282 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. BETTERIDGE AND GOAD Analyst Vol. 106 Sakamoto I. Hayashi N. and Furubayashi B. J . Phys. Colloq. 1979 2 39. Schilling J. W. J . Chem. Educ. 1979 56 A104. Sutton C. New Sci. 1979 82 540. Traexler G. Beckman Rep. 1979 15. Van den Bosch P. L. M. Janssen M. H. C. and Buck H. M. Mikrochim. Ada 1979 2 85. Vickrey T. M. and Eue W. J . Autom. Chem. 1979 1 198. Von Meerwall E. Burgan R. D. and Ferguson R. D. J . Magn. Reson. 1979 34 339. Watanabe,. F. Kagaku Sochi 1979 21 120. Watanabe K. Kagaku Kogaku 1979 43 512. Yamaguchi S. and Kusuyama T. Fresenius 2. Anal. Chem. 1979 295 256. Gran G. Analyst 1952 77 661. Johansson A. Analyst 1970 95 535. Pehrsson L. Ingman F. and Johansson A. Talanta 1976 23 769. Pehrsson L. Ingman F. and Johansson A. Talanta 1976 23 781. Osbourne A. “Introduction to Microcomputers,” Adam Osborne and Asso$ates California 1975. Aspinall D. and Dagless E. L. Editors “Introduction to Microprocessors Pitman London 1976. Barron I. and Curnow R. “The Future with Microelectronics,” Frances Pinter (Publishers), Laurie P. “The Micro Revolution,” Futura Publications London 1980. Canick A. “Computers and Instrumentation,” Heyden London 1979. London 1979. Received June 27th 1980 Accepted October 31st 198

ISSN:0003-2654    

DOI:10.1039/AN9810600257

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Analyst, March, 1981, Vol. 106, *p. 283-287 283 Computer4 inked Atom ic-a bsorption Spectroscopy: a Streamlined Approach to Environmental Analysis Mary R. Harris and N. W. Lepp Department of Biology, Liverflool Polytechnic, Byrom Street, Liverpool, L3 3A F An analytical system using two single-beam atomic-absorption spectrophoto- meters linked to a microcomputer has been developed to process efficiently large numbers of plant and soil samples for various metal analyses. The rationale underlying the development of the system and details of the inter- facing and data handling techniques are given. Keywords : Computer-linked atomic-absorption spectroscopy ; environmental analysis ; plant and soil analysis ; metal determination At present, particularly within the field of environmental monitoring of potentially toxic metals, atomic-absorption spectroscopy is widely applied, largely owing to its flexibility, ease of operation, reliability and relative cheapness.1 Modern instruments are characterised by their high degree of automation and simplicity in terms of calibration and data output.However, even with these considerable advantages, the determination of some metals, frequently at low levels for routine flame analysis, in large numbers of samples, can be extremely time consuming. In this laboratory, a 6-year research programme has been initiated to evaulate the effective- ness of different soil covering techniques in reducing trace metal uptake into soft fruit and vegetables from underlying metal-contaminated soil.2 Over this period and as a monitoring exercise, large numbers of samples will be generated and single-season workloads allow for 30000 individual metal analyses alone.Considered in the context of a biological exercise, where numerous other plant and soil parameters require examination, it will be appreciated that the analytical system utilised must be extremely efficient and relatively operator independent during bulk sample processing. Dual-channel atomic-absorption spectrophotometers are available which, under certain circumstances, effectively double the efficiency of traditional single-element systems. An alternative system is described here that utilises two individual instruments, both of which transmit concentration data to a microcomputer for data storage and processing.Experimental and Results The analytical system comprises two Pye Unicam SP9 single-beam atomic-absorption spectrophotometers. Each is served by its own SP4-01 automatic sample changer (capable of supplying 100 samples in a single run and of limitless capacity if continuously provided with sample racks) and its own SP9 computer. Both instruments have a four-lamp turret accessory and one instrument is fitted with a device to give alpha-numeric printout. Both SP9s are linked, by means of teletype cables, to a single input port on an ITT 20/20 (48KB) microcomputer which is served by a VDU and Centronics 779 printer in addition to a disk operating system consisting of two disk drive units. The instrument - microcomputer interface is made by means of a communications interface board utilising an RS232 standard.In order to permit the use of a BASIC language program, as opposed to one operating in machine code, to classify and store incoming data from the two atomic-absorption spectro- photometers, the software of the SP9 computers (ADPls) was modified. The two ADPls are linked by a two-part cable. One lead goes from an output port of the “master” ADPl to an input port of the “slave” ADPl ; this is the master’s busy line and the slave’s status line. An identical line goes from slave to master. During operation, if the master wishes to transmit a line of data to the teletype output, it sets its busy line high and checks its status line. This has to be set low before the master may proceed and ensures that the slave is not busy.Once the master has transmitted and set its busy line low, following the same logic the slave may proceed. The result is that the machines carry out each determination and transmit data in a sequential fashion. To avoid284 HARRIS AND LEPP : COMPUTER-LINKED AAS Analyst, Vol. 106 the event of both ADPls wishing to set their busy lines simultaneously, the slave, if it finds the master busy, will lower its own busy line for approximately 0.2 s to allow the master to gain control. Fig. 1 illustrates the relationship between the two ADPls and the procedures they must follow in order to transmit a line of data. Fig. 1. Flow chart showing procedures followed by master and slave ADPls before data transmission to the microcomputer can occur. An additional modification to the software of both the master and slave has the result that any line of data arriving at the microcomputer from the master has the sample number prefixed by the digit 1 and similarly from the slave by a 2.It is immediately obvious, therefore, on inspection of the data displayed on the VDU, which data refer to which instru- ment and hence to a particular element. As lead and cadmium, copper and zinc and nickel and chromium are treated as pairs, with the first always carried out by the master, no con- fusion over the determination in hand ever arises. The normal functions carried out by the SP9 computer, for example, in terms of automatic calibration, curve correction, presentation of data and its control over the operation of accessories such as the automatic sample changer and printer, are unchanged.During normal operation of the system for the determination of two elements in, for example, lettuce leaf digests, the following procedures are followed: (a) Each instrument is adjusted manually to carry out the determination of one of the pairs of elements. These operations include the selection and alignment of the appropriate hollow-cathode lamp, selection of wavelength, lamp current, band pass, burner height and fuel rates. The SP9 computer is programmed to accommodate the analysis that is to follow. The values of various parameters are keyed in: inte- gration time, number and concentration of standards and initial sample number. (b) Blank, standard and sample solutions are loaded on to each autosampler according to a specific format.With plant digests, a modification of the wet-oxidation method described by Williamss is used and the resulting 25 ml of digestate is shared between the two autosamplers, which use 12.8-ml Pyrex tubes.March, 1981 APPROACH TO ENVIRONMENTAL ANALYSIS 285 (c) The two atomic-absorption spectrophotometers are set for automatic calibration and continuous operation. (d) The ITT 20/20 microcomputer is switched on and programmed to run in floating point BASIC language. A simple, two-step, immediate command sets the baud rate of the microcomputer to 110 (as opposed to 300), thus making compatible the out- putting (atomic-absorption spectrophotometers) and receiving (ITT) devices. A data loading program is selected from disc and run. The program initially asks for the number of data lines that are expected to arrive.Once this information has been typed in, the two atomic-absorption spectrophotometers may be set to run. This stage marks the end of the manual handling of the system. From this point the system operates entirely independently of laboratory personnel. As each line of data is transmitted to the microcomputer, it is stored within the monitor as a string variable of the form A$ (I). On completion of the batch analysis, routines within the data loading program submit the stored data to a floppy disk under a specific file name. (g) At this stage, the various instrumental parameters appropriate to the next element pair may be manually selected. The data loading program is re-run to store informa- tion under a new file name and the procedures outlined above are simply repeated until the concentration of all elements of interest has been determined for the particular batch of samples.Once stored on disk, manipulation of sample concentration data can be carried out using various computer routines. In the present case, an initial visual inspection of the file is made, using the hard-copy printout, to check for curvature, calibration and the presence of spurious sample readouts. Using a second calculator program, all readouts pertaining to sample concentrations only are assigned to a new string array from which, using conventional commands for the conversion of strings to numerical values, concentration data can be corrected for blank values, dilution and sample dry mass parameters.A formating section presents individual sample concentrations (pg 8-1 dry mass) in two columns relating to the element pair represented. A third and final program is used to collect all concentration data from each element pair file, first to store a comprehensive record of concentration data in numerical form (for statistical work) and finally to format the results as a complete table of coded samples together with the concentration of each element determined. Tables 1-111 give examples of the type of readout that can be obtained, if required, at each stage of the proceedings. (e) (f) This file forms the basis of all subsequent data analysis. Conclusion The system described has a number of important advantages : ( a ) Efficiency of operation for the detection of, routinely, six elements for large numbers (typically of the order of 100-150 per batch) of soil extracts and plant digests. ( b ) Flexibility in pairing of elements to be analysed and in the numbers of pairs (and, if required, single-element operation) that may be included.An important considera- tion is that analyses are often carried out at low concentrations (typically in the range for lead and cadmium of 1.00-0.05, nickel and chromium < 0.50 and copper and zinc < 2.00 pg ml-l), and as the operating conditions of each instrument are set independently to optimise sensitivity and reduce noise for the determination in hand, no compromise in these conditions is required as may occur with dual-channel systems. Real savings in analytical and data handling time.Under ideal conditions a typical batch of 100 samples can be analysed for six elements in approximately 2.5 h. This estimate is based on a 10-min interval for the manual setting of instrument para- meters together with a 40-min autocalibration and analytical cycle using integration times of 10 s. An additional 0.5 h may be required for loading samples on to racks for each autosampler. On a purely manual basis, it is estimated that with a 5-min period for parameter settings, 5-min calibration time, 5-s handling and 10-s integra- tion time per sample, the processing and analysis time would be 3.5 h. An additional 1 h would be required for loading the resulting 600 data points into a microprocessor for data processing. (c)286 0.01 0.02 0.62 3.26 0.52 2.65 0.63 2.79 \ 0:74 3.59 HARRIS AND LEPP : COMPUTER-LINKED AAS Analyst, Vol.106 TABLE I (SAMPLE NO. PRECEDED BY 2) DETERMINATIONS IN LETTUCE LEAF DIGESTS FORMAT O F INITIAL DATA FILE FOR CU (SAMPLE NO. PRECEDED BY 1) AND ZN DURING AUTOMATIC PROCESSING Key to right identifies solution type. AAS denotes instruments are in flame absorption rather than flame emission mode. Element No. code mode pg ml-l Sample Sample Instrument concentration/ Key “real” sample concentrations ’ (PET ml-l) 1997 BL AAS 0.0000 o.ooo }blank value 2997 BL AAS 1998 S1 AAS 0.084 o.049 }standard 1 absorbance 2998 S1 AAS 1998 CON AAS 1.00 standard 1 concentration 2998 CON AAS 2.00 } (pgml-1) 0.164 o. } standard 2 absorbance 1999 s 2 AAS 2999 S2 AAS 1999 2999 1000 2000 1001 2001 1002 2002 1003 2003 1100 2 100 CON CON CON CON CON CON CON CON CON CON CON CON AAS AAS AAS AAS AAS AAS AAS AAS AAS AAS AAS AAS TABLE I1 FORMAT OF SECOND DATA FILE SHOWING CORRECTED CONCENTRATIONS OF CU AND Z N I N SAMPLES Sample No.1001 1002 1003 1004 1005 1006 1007 Cu concentration1 pg g-’ dry mass 15.25 12.75 15.50 15.75 15.25 14.25 15.75 Sample No. 2001 2002 2003 2004 2005 2006 2007 Zn concentration/ pg g-l dry mass 81 .OO 65.25 69.25 69.00 65.25 77.75 69.60 1100 18.25 2100 89.25 Two of the most significant advantages, however, are the release of often highly skilled personnel from repetitive tasks and the avoidance of transcript errors during data handling. (d) General applicability to computer systems that possess an RS232 standard interface and to personnel without machine language expertise.(e) Cost : although the system described above (two atomic-absorption spectrophoto- meters, two sample changers, 48KB computer, VDU, dual disk drives and printer) cost 12% more (1979 basis) than a comparable dual-channel atomic-absorption spectro- photometer with sample changer alone, the inherent flexibility of this system justifiesMarch, 1981 Plot A01-1 . . A01-2 . . A09-1 . . A09-2 . . A23-1 . . A23-2 . . A32-1 . . A32-2 .. P31-1 .. P31-2 .. APPROACH TO ENVIRONMENTAL ANALYSIS TABLE 111 FINAL RESULTS FOR CODED SAMPLES r- Pb . . 3.75 . . 4.75 . . 3.50 . . 2.25 . . 2.25 . . 2.00 . . 2.50 . , 1.25 Element concentration/pg g-1 dry mass Cd c u Zn Ni 2.63 15.25 81.00 1.13 1.88 12.75 65.25 0.63 1.62 15.50 69.25 1.62 1.63 15.75 69.00 1.63 0.88 15.25 65.25 1.38 2.38 14.25 77.75 0.88 2.13 15.75 69.50 1.63 1.12 15.50 68.00 0.87 A .. 26.53 26.69 17.b 87:OO 7.56 . . 22.98 24.10 18.25 89.25 7.62 287 7 Cr 1.75 1.50 1.50 1.50 1.00 2.00 1.25 1.25 5.86 5.99 its adoption. In addition, in order to obtain the reductions in data processing time a microcomputer would also be required for a dual-channel instrument, which would produce a reversal in the cost estimates. In conclusion, experience gained with this system to date establishes the concept of computer-linked atomic-absorption spectroscopy as a major advance in automating many of the time-consuming aspects of analytical procedures, with concurrent savings in time and cost. This work was carried out on behalf pf, and funded by, the Department of the Environ- ment. The authors thank the staff of Pye Unicam (UK), particularly those in the Develop- ment and Technical Departments. References 1. Kirkbright, G. F., in Pam, R., Editor, “Elemental Analysis of Biological Materials by Atomic Absorption Spectroscopy,” in “International Symposium on Nuclear Activation Techniques in the Life Sciences.” Proceedings of the International Atomic Energy Agency Conference, Inter- national Atomic Energy Agency, Vienna, 1978, pp. 141-165. Lepp, N. W., and Harris, M. R., “A Strategy for Evaluation of Soil Covering Techniques to Reduce Trace Metal Uptake by Soft Fruits and Vegetables,” in “Reclamation of Contaminated Land,” Proceedings of Society of Chemical Industry Conference, Eastbourne, Society of Chemical Industry, London, 1979, Section C7. 2. 3. Williams, E. V., J . Food Technol., 1978, 13, 367. Received August Bth, 1980 Accepted October 13th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9810600283

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

288 Analyst, March, 1981, Vol. 106, $9. 288-298 A Novel Method of Wavelength Modulation for Atomic Spectrometry-Some Preliminary Experiments R. G. Michel," J. Sneddon, J. K. Hunter and J. M. Ottaway and G. S. Fell Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow, G1 1XL Department of Clinical Biochemistry, Royal Infirmary, Glasgow, G4 OSF The method of wavelength modulation most commonly used in atomic spectrometry employs an oscillating refractor plate driven by a scanning motor. A new mechanical arrangement is described, which involves a rotating quartz mechanical chopper. This quartz chopper is composed of four quad- rants of different thickness, which refract radiation by different degrees. The rotating quartz chopper is placed inside the monochromator in a position similar to that used for oscillating refractor plates.Efficient wavelength modulation is achieved. However, the rotating quartz chopper has the advantage that a good square-wave modulation function is obtained. Keywords : Square-wave wavelength moddation ; atomic spectrometry Wavelength modulation is now well established as a means of background correction in atomic spectrometry. It has been applied particularly successfully to atomic emission (AE) in electrothermal atomisers,1s2 flame^^-^ and plasma^,^^^ to continuum source excited atomic absorption (AAC)9,10 and, to a lesser extent, to continuum source excited atomic fluorescence (AFC).11,12 The primary advantage of wavelength modulation lies in its ability to discrimin- ate against a continuum whilst measuring atomic line spectra.This leads to improvements in accuracy of atomic spectrometric measurements. The continuum may be atom cell back- ground emission (AE, AAC, AFC), scatter of excitation source radiation (AFC, AAC), matrix emission (AE, AAC, AFC), excitation source continuum viewed directly (AAC) and non- analyte, broad-band, absorption (AAC) or fluorescence (AFC) . The discrimination against continuum background can lead to improvements in signal to noise ratio owing to concomitant discrimination against low-frequency fluctuation noise.13 Wavelength modulation using the zero-crossing technique5 can also assist in correction for spectral interferences. The method of wavelength modulation most commonly used in atomic spectrometry employs an oscillating refractor plate.The principle of this was probably first described by M~Wi1liam.l~ It is current practice to cause the refractor plate to oscillate about the vertical axis using commercially available scanning motors. Most researchers have used a sinusoidal waveform to drive the scanning motor because this is the most straightforward to set up experi- mentally. In 1977 Koirtyohann et aZ.15 and O'Haver et aZ.13 presented experimental and theoretical results, respectively, that demonstrate that square-wave oscillation of the refractor plate improves the ultimate signal to noise ratio of the measurement by a factor of about 1.8. With scanning motors, square-wave oscillation is more difficult to achieve than sine-wave oscillation a t frequencies acceptable for atomic spectrometry.However, commercial equipment is availablel5 that incorporates suffici- ent control over the position of the plate to achieve good square-wave modulation at about 17 Hz. In this paper we describe a mechanical arrangement that achieves wavelength modulation and that results in the wavelength being scanned with a square waveform at about 20 Hz, with potential for higher frequencies. The arrangement involves a rotating quartz mechanical chopper composed of four quadrants of different thickness, which refract radiation by different degrees. The rotating quartz chopper is placed inside the monochromator in a position similar to that used for an oscillating refractor plate and is a more straightforward and cheaper means It results in the wavelength being scanned sinusoidally.* Present address : Department of Chemistry, University of Connecticut, Storrs, Conn. 06268, USA.MICHEL, SNEDDON, HUNTER, OTTAWAY AND FELL 289 of wavelength modulation than an oscillating refractor plate. This is a result of the mechanical arrangement being no more complex than traditional rotating glass choppers used for intensity modulation in spectrometric instruments. Moreover, no complex electronic control is required to drive the chopper. The experiments described here were involved with flame atomic- fluorescence spectrometry. However, there is no reason why the system should not be equally successfully applied to atomic-emission and atomic-absorption measurements. Experimental Instrumentation The rotating quartz mechanical chopper was incorporated into a flame atomic-fluorescence instrument built at the University of Strathclyde and which has been described elsewhere.16~17 The line sources discussed in references 16 and 17 were not used.The 300-W xenon arc, normally used at low power for background correction, was operated here at almost maximum power to excite atomic-fluorescence signals in the flame. A schematic diagram of the relevant parts of the instrument is shown in Fig. 1. Fig. 2 shows the quartz mechanical cho per. The circular quartz chopper blade was con- tific Ltd., Heme1 Hempstead, Hertfordshire, UK). Each quadrant was cut from one of three circular discs 50 mm in diameter and 1, 2.5 or 4 mm thick. Two quadrants of the resulting chopper were 2.5 mm thick.The third and fourth quadrants were 1 and 4 mm thick and they were arranged as indicated in Fig. 2. This configuration allows primarily for wavelength modulation in the “2F” mode, although a “1F” mode is possible. The 2F mode allows for background measurement/correction on both sides of the atomic line and is achieved by passing the atomic line wavelength through the exit slit when the 2.5-mm quadrants are in the optical beam. The 1F mode allows for background measurement/correction on one side of the atomic line only and is obtained with the above system by using either the 1- or 4-mm quadrants for the atomic line. In each instance the three other quadrants will correspond to wavelengths on the same side of the atomic line, at either higher or lower wavelengths.All of the possible forms of modulation are discussed later. Each of the four quadrants was drilled in the position indicated in Fig. 2 and bolted and cemented to the hub of a 33 mm diameter gear wheel (hub diameter 20 mm). The bolts served primarily to locate the quartz accurately during assembly. structed of four quadrants of optically poli 5 hed quartz (Vitreosil, supplied by Jencons Scien- Photon counter Mirror TL Reference I I I i Spherical Double monochromator U- mirror Fig. 1. Atomic-fluorescence instrument with continuum source For details of instrumentation excitation and wavelength modulation. not described in the text, see reference 16. A second identical gear wheel, driven by a synchronous motor, was meshed with the chopper assembly.This arrangement rotated the chopper at 10 rev s-l to give a modulation frequency of 20 Hz in the 2F mode. Direct drive by the motor was not possible because the bulk of the290 MICHEL et d. : WAVELENGTH MODULATION Analyst, Vol. 106 motor would have prevented passage of the light beam through the chopper blades. In the double monochromator used here the image of the middle slit was passed through the top half of the chopper blade and aligned along the vertical diameter. The second gear wheel had slots cut in its body as indicated in Fig. 2. A miniature infrared light-emitting diode (LED, 306-077, R.S. Components Ltd., London) was mounted in the main support bracket at the back of the gear and slots. The infrared radiation was detected by a phototransistor (306-083, R.S.Components Ltd.), which was placed on a post in front of the slots. This arrangement served to provide a reference signal for the synchronous detection system of the photon counter (Fig. 1) that was used to process the signals from the photo- multiplier tube. An infrared reference system was preferable to a conventional tungsten bulb system in order to exclude stray light from the monochromator. The photomultiplier tube (9789QB, EMI, Hayes, Middlesex) had a response that fell rapidly a t wavelengths longer than 650 nm. Frovision was necessary to allow'variation of the angle of incidence of the light beam at the chopper. This was achieved by pivoting the assembly about the vertical diameter of the chopper blade. After the angle of incidence had been set, the assembly was tightened down to the base of the monochromator with a screw running in a radial slot located as indicated in Fig.2. Fig. 1 shows the location of the chopper assembly inside the double monochromator of the instrument. It is placed just after the middle slit, which is the entrance slit of the second monochromator in the double monochromator. Reagents All reagents were of the highest purity available. High-purity de-ionised water was used for the preparation of all solutions. Metal stock solutions were prepared by dissolving a known amount of spectrographically pure metal usually in 10 ml of 11 M hydrochloric acid. All solutions for analysis were adjusted to be 0.04 M in hydrochloric acid for maximum stability of low concentrations of metal in solution.The preparation of blood samples for copper determination was carried out as described by Peaston18 for both atomic-fluorescence and atomic-absorption measurements. Instrumental Conditions publications,16~l7 unless stated otherwise. All instrument settings and operating conditions were the same as those reported in previous Design Principles of the Rotating Quartz Chopper Elsewhere,3~~~ the basic principles of the use of a refractor plate to obtain a rapid scan over a small wavelength interval have been described. Until the present, a wavelength scan has normally been achieved by placing an oscillating refractor plate in the light path, inside the monochromator, either just after the entrance slit or just before the exit slit. The light beam, in passing through the plate, is displaced laterally by an amount, d , given by the relationship3 12-1 d " t .( n > .. .. where t mm = thickness of the plate, a rad = angle of incidence of the light beam a t the refractor plate (see Fig. I) and n = refractive index of the quartz plate. The rotating quartz chopper operates on the same principle as the oscillating refractor plate except that, in order to vary d and hence the wavelength, the thickness, t, is varied rather than the angle of incidence, a. Square wave As the quartz chopper rotates, each plate in turn refracts the light beam to an extent depend- ing on the thickness of each plate. The boundary between each plate is sharply defined and provides the basis for the resulting square-wave variation of wavelength with respect to time.However, the change in direction of the refracted beam, which begins to take place as the beam begins to irradiate the next plate in the sequence, occurs gradually in a time related to the width of the light beam and the rotational velocity of the chopper blade. At a constantLight-beam path 1 Quadrants. 1-4 Holes for mounting ' bolts . . Synchronous motor Gear wheels / Support / bracket Infrared LED 'and Dhototransis Reference slots ;tor I Radial slot Pivot point Fig. 2. Quartz mechanical chopper assembly. Quadrant thicknesses were as follows: 1 and 2, 2.6 mm [tl and t, in equation (2)]; 3, 1 mm [t3 in equation (2)]; and 4, 4 mm [tr in equation (2)]. [to face page ago( a ) 2F mode (b) 1F mode (case 2, see text) Fig. 3. Photographs of square wave using rotating quartz chopper.The light source was a lead hollow-cathode lamp observed directly. The monochromator was an Cchelle (Spectrametrics Inc., Andover, Mass., USA). The signals were taken directly from the photomultiplier tube and displayed on an oscilloscope.*O The oscilloscope axes are indicated a t 100 mV cm-1 vertical and 10 ms cm-l horizontal.March, 1981 FOR ATOMIC SPECTROMETRY 291 rotational velocity the ratio of one quarter of the circumference of the chopper to the width of the light beam determines the proportion of the square wave taken up by the rise and fall times, i.e., for a chopper of fixed size the width of the light beam determines the rise and fall times. The width of the light beam is initially the width of the entrance slit of the monochromator.By the time it reaches the chopper it has diverged by about 2-3 mm, depending on the distance between the entrance slit and the chopper. It is therefore desirable to place the chopper as close as possible to the entrance slit in order to obtain a well defined square wave. The size of the chopper assembly must be minimised to facilitate a close approach to the entrance slit. Some spurious reflections of light a t the junctions between plates were noticeable when observing the square wave on an oscilloscope. These were rejected by reducing the duty cycle of the detection system by about 20%. This facility was available on the synchronous detector of the photon counter employed and it allows the observation of the first and last 10% of each chopper quadrant to be rejected.Recent trials20 have demonstrated that most of the spurious reflections can be reduced by applying matt black paint to the edges of each plate a t each junction. Fig. 3 shows photographs of the square-wave modulation in the 2F mode (a) and in a pseudo-1F mode (b) (see case 2 below). These photographs were obtained fron a more recent version of the rotating optical chopper, which was not used for the remaining results in this paper but which will be described at a later date.20 Use of a hollow-cathode lamp indicates clearly the observation of the atomic line (noisy portions) and background on either side of the line (lower noise). The photographs in Fig. 3 demonstrate the excellent square wave obtainable after reduction of spurious reflections by application of optically black paint to plate edges.The modulation frequency in Fig. 3(a) was 40 Hz. Modulation interval The arrangement of thicknesses of the four plates shown in Fig. 2 was designed to give wave- length modulation in the 2F mode. The two 2.5-mm plates allow the atomic line wavelength, h, determined by the grating position, to be directed a t the exit slit of the monochromator. The 1.0-mm plate refracts light less than the 2.5-mm plate and therefore irradiates the exit slit with radiation of wavelengths displaced to one side of the atomic line. The 4-mm plate refracts light more than the 2.5-mm plates and light from the opposite side of the atomic line is detected. The magnitude of the displacements is determined by the difference in thickness between the 1.0 and the 4.0-mm plates and by the angle of incidence of radiation falling on the plates.This means that for plates of fixed thickness the total displacement (modulation interval, AA), can be varied by varying the angle of incidence. The displacement of the light beam can be determined from equation (l), i.e., for the thickest plate (t, mm) and the thinnest plate (t, mm) the displacement, d, is given by AA = d = (t4-t,) a (Q) . . For the rotating quartz chopper described in Fig. 2, t, was 1.0 mm. This was the minimum thickness possible that had acceptable mechanical strength. The plate with t , = 4 mm was chosen on the basis of the displacement required for the monochr~mator~~ on our instrument. This was a 220-mm focal length, f/4, double monochromator, which, for atomic-fluorescence measurements, was used with a spectral band pass, 8A, of 0.5 nm a t slit widths of 0.25 mm (the entrance and exit slits were 0.25 mm, the middle slit was 1.25 mm wide).This corresponds to a minimum total displacement of 0.5 mm for a peak-to-peak modulation interval,13 AA = 28A, of 1 nm for measurements in the 2F mode. From equation (2), t,-t, is 2.86 mm a t an angle of incidence, a, of 30°, displacement, d , of 0.5 mm and refractive index, n, of 1.5. A convenient plate thickness, t,, of 4 mm was chosen for a t,--t, of 3 mm. At a = 30" this gives a slightly greater modulation interval than 28A and larger values of AX are possible by increasing the angle of incidence. This was considered necessary because reports13 have indicated that AA may need to be up to 2.56h for maximum signal to noise ratio.2F mode The square modulation waveform in the 2F mode using the rotating quartz chopper can be292 MICHEL et al. : WAVELENGTH MODULATION Analyst, Vol. 106 defined in terms similar to the three-step waveform described by O'Haver et al.13 for the oscil- lating refractor plate. When the chopper rotates through one complete revolution from 0 to 2 T radians the wavelength changes are shown in Fig. 4, A, and are as follows: the atomic line at i is observed through the two 2.5-mm plates ( A = x at 0 to n/2 and T to 3n/2) and background measurements are made using the 1-mm plate (A = i-AA/2 at n/2 to 72) and the 4-mm plate ( A = h + AA/2 at 3 ~ / 2 to 2 T). This is the normal form of modulation for which the system described was designed and allows correction at each side of the atomic line wavelength and maximum signal to noise ratio.0 x n 32 2n 2 2 Rotation of chopper, rad Fig. 4. Change in wavelength with plate thickness. Chopper rotating through 27r rad. For further explana- tion, see text. Numbers in parentheses refer to thickness of plate (mm) during the relevant part of the modulation cycle. A, 3-step, 2F modulation; B and C, 2-step, 1F modulation; and D and E, 3-step, 1F modulation. 1F mode-case 1 The 1F mode can be obtained by constructing a new chopper blade with only two plate thicknesses. Fig. 4, B, illustrates this for such a two-step chopper constructed of four quad- rants alternating between 1- and 2.5-mm thickness and the grating tuned so that h is observed through the 2.5-mm plates ( A = h at 0 to n/2 and T to 3 ~ / 2 ) and background measurements are made using the 1-mm plates ( A = A - AA at 7r/2 to n and 3 ~ / 2 to 2 n).Note that AA in Fig. 4, B and C, is only half the modulation interval seen in Fig. 4, A. Fig. 4, C, shows the same two- step chopper with the grating tuned so that h is detected through the 1-mm plates ( A = h, at n/2 to v and 3 ~ / 2 to 2 7r) and background measurements are made using the 2.5-mm plates ( A = 1 + Ah at 0 to n/2 and n to 3n/2).March, 1981 FOR ATOMIC SPECTROMETRY 293 1F mode-case 2 It is inconvenient to change chopper blades in order to change from the 2F to the true 1F mode. However, it is possible to make pseudo-lF measurements using the 2F (three-step) chopper.This is illustrated in Fig. 4, D, where the atomic line is observed through the 4-mm plate (A = Ti at 3n/2 to 2 n) and background measurements made using the 1-mm plate (A = A-AA at n/2 to v) and the 2.5-mm plates (A = X-AA/Z at 0 to 4 2 and n to 3 ~ / 2 ) . In Fig. 4, E, the atomic line is observed through the 1-mm plate (A = at n/2 to n) and back- ground measurements are made at the opposite side of the line to 4, D, using the 2.5-mm plates ( A = + Ah a t 3n/2 to 2 n). It is clear, from Fig. 4, D and E, and Fig. 3(b), that the atomic line is only observed for half of the time that it is observed in the 2F mode of Fig. 4, A. This implies a decrease in signal to noise ratio, at the detection limit, of a factor of 2 compared with the 2F mode.This would not occur if a true two-step chopper were to be used as indicated in Fig. 4, B and C. However, in most circumstances the pseudo-1F mode may be satisfactory relative to the inconvenience of changing chopper blades. The synchronous detector must be sensitive to a signal at half the chopping frequency, or the reference signal could be electronically halved in frequency in order to detect the pseudo-1F signal. The photon counter with synchronous detector used in our experiments registered half the count rate in the 1F mode compared with the 2F mode because the signal channel was not tuned or filtered as would be the case when using an analogue lock-in amplifier. The photon counter therefore presented no problems for detecting 1F signals by retuning the monochromator when using the 2F (three-step) chopper blade.A ngle of incidence It is interesting that the angle of incidence of the light beam on the rotating quartz chopper is constant throughout the modulation cycle. This minimises changes in reflection losses during the modulation cycle and contrasts with an oscillating refractor plate where the reflection losses change because the angle of incidence is changing during the modulation cycle. However, this is not a real advantage for the rotating glass chopper because the change in reflection losses with angle of incidence is negligible21 during the conventional maximum angle of oscillation of between 0 and &15". Results and Discussion + AA/2 at 0 to n/2 and n to 3n/2) and the 4-mm plate (A = The instrument used for these experiments was designed primarily for atomic-fluorescence measurements using line sources and background correction with a continuum source. The light sources were intensity modulated in the normal manner for a two-source background correction system.16 Here the continuum source was used alone and at maximum power to excite atomic fluorescence.Ideally, double modulation (source intensity and wavelength modulation) is required to correct for both atomic emission and scatter of source radiation in addition to flame background.1l If wavelength modulation is used without source intensity modulation then both atomic emission and atomic fluorescence of the analyte metal contribute to the total signal. A disadvantage of this approach is the likely incidence of spectral inter- ferences from structure in the flame background and from structure in matrix thermal emission in the flame.However, prior to assembling an instrument incorporating double modulation, the rotating quartz chopper was used for combined atomic-fluorescence/atomic-emission measurements. This provided an initial evaluation of the utility of the quartz chopper for background correction. Detection Limits A comparison was made between the detection limits (1 s count time, signal to noise ratio = 2) obtained using source intensity modulation alone, with detection limits obtained using wavelength modulation alone. In column 2 under Detection limit, atomic emission from the analyte does not contribute to the total signal as the source was intensity modulated.Hence, the detection limits for those elements which are sensitively determined by atomic emission (resonance wavelengths longer than 300 nm) show an improvement in detection limit in going from source intensity modulation (column 2), to wavelength modulation (column 1). At wavelengths shorter than 300 nm, the detec- tion limits are similar for both methods of modulation. These results are shown in Table I.294 MICHEL et al. : WAVELENGTH MODULATION Analyst, Vol. 106 At wavelengths shorter than 250 nm the detection limits appear worse for wavelength modulation, probably because of a loss of ultraviolet output of the xenon arc owing to lamp ageing between the two sets of results, which were obtained about 1 year apart. There is probably also a slight contribution to the poorer detection limits for wavelength modulation due to transmission losses of around 10% at the refractor plates.TABLE I ATOMIC-FLUORESCENCE DETECTION LIMITS* Excitation, 300-W continuum source ; nitrogen-separated air - acetylene flame. Detection limitlpg 1-1 Wavelength/ - Element t nm 1: 2: Zinc . . .. .. .. 213.9 195 51 Cadmium .. .. .. 228.8 44 19 Nickel .. .. .. 232.0 106 99 Cobalt .. * . .. 240.1 148 49 Iron . . .. .. .. 248.3 98 19 Gold . . .. .. .. 267.5 390 390 Manganese . . .. .. 279.5 7 4 Lead . . .. .. .. 283.3 360 350 Magnesium . . .. .. 285.2 1.3 0.7 Copper . . .. . . 324.7 7 20 Silver .. .. . . 328.1 3 4 Chromium . . .. . . 357.9 6 24 Strontium . . .. . . 460.7 6 7.6 * For definition see text-count time was 1 s.t Aqueous solution 0.04 M in hydrochloric acid. 1, Wavelength modulation, rotating chopper ; 2, source intensity modulation. The detection limits were obtained using noise figures calculated by taking the square root of the total background measured in counts per second. This assumed that the background shot noise was dominant, which has been shown previously22 to be true for source intensity modu- lated, continuum source excited flame atomic fluorescence. No attempt was made to deter- mine whether there was a low frequency, lr, noise component in the wavelength modulation situation described here. O’Haver et al.13 have presented a discussion of the predicted effects of wavelength modulation on signal to noise ratio in various atomic spectrometric situations.This is at present the subject of further measurements. Second Derivative Wavelength Scan of Atomic Line Wavelength scans of atomic lines with the rotating chopper in operation were essentially analogous to those obtained by Snelleman et aL3 for the oscillating refractor plate. An example of the spectrum for the cadmium atomic fluorescence line at 221.8 nm is shown in Fig. 5. The apparent steps in the peaks are a result of the digital-to-analogue conversion at the output of the photon counter. Inter-relationship Between Refractive Index, Angle of Incidence and Modulation Interval The refractive index of the Vitreosil quartz used in the rotating quartz chopper varies with wavelength from 1.52 a t 200 nm to 1.44 a t 600 nm, and this affects the modulation interval in a manner dictated by equation (2).The modulation interval decreases with decreasing refrac- tive index. From equation (2) it can be seen that it is possible to compensate for the change in modulation interval by varying the angle of incidence. This is equivalent to varying the amplitude of the oscillation when using an oscillating refractor plate3 and both achieve the same purpose of controlling the modulation interval. The variation of modulation interval with wavelength dictates that the angle of incidence be optimised for maximum signal whenever measurements are to be made a t different wavelengths. This is illustrated in Table 11, which lists the optimum angle of incidence that was found to beMarch, 1981 FOR ATOMIC SPECTROMETRY 296 40 > C C al m .= 20 w .- .- c - a 0 a - 10 B Wavelength scan (1 nm min-’) Fig. 6.Atomic-fluor- escence spectrum of cad- mium at 228.8 nm. Scan rate, 1 nm mm-l ; nitrogen- separated air - acetylene flame; solution concen- tration, 500 p g ml-l in 0.04 M hydrochloric acid; and spectral band pass, nominally 0.5 nm. A, Cadmium atomic fluorescence observed through 4-mm plate; B, sum of cadmium atomic fluorescence observed through each 2.5-mm plate; and C, cadmium atomic fluorescence observed through 1-mm plate. required to achieve the optimum detection limits for each metal listed in Table I. If the angle of incidence was not optimised at each wavelength but was kept constant at a value optimised in the ultraviolet range, say for zinc, then detection limits deteriorated by up to a factor of two or three for the remaining metals, The results in Table I1 were obtained by aspirating a solu- tion of each metal of interest into the flame and then adjusting the angle of incidence to maxi- mise the atomic fluorescence signal to (background)+ ratio.The angle of incidence was then measured directly in the monochromator using a protractor (error about &lo). The modula- tion intervals which resulted were calculated from the experimentally determined angles of incidence and the known refractive index at each wavelength. These results, shown in the last column of Table 11, revealed that the modulation interval had remained essentially con- stant (average displacement of 0.685 mm) throughout the wavelength range studied (213- 461 nm). The 0.685-mm displacement corresponds to a modulation interval of 2.74 6h for the 0.25-mm slit width used (spectral band pass 0.5 nm, i.e., AA = 1.37 nm).The expected modulation interval for maximum signal is 2.0 6A or greater.13 For sinusoidal modulation the maximum signal to noise ratio is predicted to be at 2.5 The figure of 2.74 6h found here was prob- ably a result of two factors. Firstly, with the rotating refractor plates in position there will be some defocusing of the image of the exit slit.3 This will tend to increase the spectral band pass, ah, and therefore the required modulation interval. Secondly, there was some departure from the ideal square wave, which was a result of the finite width of the light beam incident on the296 MICHEL et al. : WAVELENGTH MODULATION Analyst, Vol.106 refractor plates. This implies that the modulation waveform had some sine wave character, which would tend to increase the required modulation interval for accurate discrimination between signal and background. Discrimination Against Flame Background A wavelength scan between 200 and 600 nm revealed that the wavelength modulation was effectively discriminating against flame background when using the rotating quartz chopper (Fig. 6). The scan was taken with the angle of incidence at the chopper set at 40°, which is optimum only for wavelengths below 230 nm (Table 11). TABLE I1 ANGLE OF INCIDENCE REQUIRED TO ACHIEVE OPTIMUM DETECTION LIMITS LISTED IN TABLE I Element Zinc . . .. .. .. Cadmium .. .. .. Nickel .. .. .. Cobalt .. .. .. Iron .. .. .. .. Gold . . .. .. .. Manganese . . .. .. Lead . . .. .. .. Magnesium . . .. .. Copper . . .. .. Silver .. .. .. Strontium . . .. .. Modulation interval/ Wavelength/ Angle of incidence, pm of nm degrees displacement 213.9 228.8 232.0 240.1 248.3 267.5 279.5 283.3 285.2 324.7 328.1 460.7 40 40 40 41 41 41 41 41 43 42 43 44 685 68 1 674 686 684 676 673 672 704 682 697 703 Average: 685 I I 200 300 400 i 500 Wavelengthhm 4 - ! 600 Fig. 6. Residual flame background after wavelength modulation. Spectrum obtained by scanning monochromator wavelength and aspirating water into a nitrogen-separated air - acetylene flame whilst the rotating quartz chopper was in operation. The angle of incidence was 40".March, I981 FOR ATOMIC SPECTROMETRY 297 The magnitude of the flame background, when measured using d.c.detection, varies between about 200 counts s-l at 230 nm and about loo00 counts s-l at around 350 11m.l' The maxi- mum offset in Fig. 6 is about 40 counts s-l at 350 nm, i.e., less than 0.5% of the total flame background and, therefore, will cause negligible inaccuracies in measurements with real samples. This small offset may have been due to disproportionate light losses in the 4-mm refractor plate relative to the l-mm plate, leading to a slightly inaccurate average measurement of the flame background. The large noise peaks in the spectrum occur in regions of the flame background which contain line structure such as the OH bands at around 310 nm. The small offset was reproducible and was not, therefore, due to low-frequency drift.Determination of Copper in Blood Serum Table I11 shows the results for the determination of copper in blood serum using both the wavelengt h-modulated atomic-fluorescence/atomic-emission instrument described here and a standard flame atomic-absorption instrument. The method used to prepare samples, stand- ards, etc., was identical with that described by Peaston.l* The blood samples were obtained from the Royal Infirmary Glasgow, and were from a non-normal sample of people (hospital patients). The results of the analyses using both methods gave a correlation coefficient between the two sets of results that showed good agreement (Table 111). Further comparisons with more diverse methods of copper determination are required in order to characterise fully the accuracy of the atomic-fluorescence/atomic-emission, wavelength-modulation method, relative to other techniques. However, these results do indicate that the wavelength-modulation technique described here is giving satisfactory corrections for flame background and scatter of excitation source radiation as otherwise a positive bias would have been obtained for the atomic- fluorescence/atomic emission method.Research is continuing to obtain further support for these results. TABLE I11 DETERMINATION OF COPPER IN BLOOD SERUM USING TWO TECHNIQUES Flame atomic-fluorescence spectrometry (AFS) with wavelength modulation using a nitrogen-separated air - acetylene flame; flame atomic-absorption spectrometry (AAS) on a Perkin-Elmer 403 instrument with a standard air - acetylene flame.Methodology as in reference 18. Copper contentlpg ml-l Sample No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 I 1 AAS AFS 1.57 1.61 1.32 1.32 1.14 1.21 0.88 0.86 1.27 1.25 1.22 1.10 1.67 1.72 1.04 0.97 1.67 1.81 1.05 1.04 1.26 1.31 1.18 1.18 1.11 1.12 1.45 1.42 Sample No. 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Copper contentlpg ml-l AAS AFS 1.89 1.72 1.20 1.19 1.41 1.30 1.32 1.28 1.91 1.95 1.21 1.19 0.98 0.96 0.95 0.91 1.09 1.15 1.49 1.56 1.56 1.59 1.27 1.23 1.95 2.02 1.46 1.46 A 7 7 Mean copper concentration/pg ml-l. . .. .. . . = 1.340(AAS); 1.337(AFS). Standard deviationlpg ml-l . . . . .. . . . . = 0.291(AAS); 0.31O(AFS). Correlation coefficient between the results of each technique Regression equation . . .. .. .. .. .. = A =mF+C (A =AAS result; Gradient of regression equation .. . . .. . . = 1.041. Range/pg ml-' . . .. .. .. . . .. . . = 0.88-2.02. = 0.979, P< 0.05 at 95% confidence. F = AFS result). Intercept of regression equationlpg ml-l . . . . .. - - -0.058.298 MICHEL, SNEDDON, HUNTER, OTTAWAY AND FELL Conclusions The rotating quartz chopper described here gives excellent square-wave modulation, and background correction with acceptable accuracy, both for the determination of a metal in a real sample (copper in blood serum) and for the discrimination against flame background. Further research is in progress to resolve the small inaccuracies (less than 0.5%) that still exist after background correction. The potential of this rotating quartz chopper is in its possible development for higher fre- quencies of square-wave modulation.This is limited only by the maximum possible safe rotational speed of such a glass chopper and the physical size of the chopper blades. Larger blades will allow a better square wave to be obtained in a manner comparable to the use of conventional mechanical choppers. These considerations are relatively easy to resolve and it appears feasible that modulation frequencies of around 200 Hz could be obtained by careful design of the chopper and its physical location in the monochromator. The mechanical arrangement is no more complex than traditional rotating glass choppers used for intensity modulation in spectrometric measurements and no electronic control is re- quired. (An oscillating refractor plate does require electronic control of its scanning motor in order to achieve square-wave modulation.) The rotating optical chopper is equally applicable to atomic emission, fluorescence and absorption and is, in this respect, comparable to an oscillating refractor plate.The authors acknowledge the support of the Scottish Home and Health Department for the purchase of the major items of equipment used in this project and for the award of a Post- doctoral Fellowship (to support R.G.M.). They also thank the Eastern District, Glasgow Health Board, for a maintenance grant in support of J.S. 1 . 2. 3. 4. 5. 6 . 7 . 8 . 9. 10. 1 1 . 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References Epstein, M. S., Rains, T. C., and O’Haver, T. C., Appl. Spectrosc., 1976, 30, 324. Hutton, R. C., Ottaway, J . M., Epstein, M. S., and Rains, T. C., Analyst, 1977, 102, 658. Snelleman, W., Rains, T. C., Yee, K. W., Cook, H. D., and Menis, O., Anal. Chem., 1970, 42, 394. Rains, T . C., and Menis, O., Anal. Lett., 1974, 7, 715. Epstein, M. S . , and O’Haver, T. C., Spectrochim. Acta, 1975, 30B, 135. Sydor, R. J., and Hieftje, G. M., Anal. Chem., 1976, 48, 535. Kawaguchi, H., Okada, M., Ito, T., andMituike, A., Anal. Chim. Acta, 1977, 95, 152. Rose, O., Jr., Mincey, D. W., Yacynych, A. M., Heineman, W. R., and Caruso, J. A., Analyst, 1976, Snelleman, W., Spectrochim. Acta, 1968, 23B, 403. Zander, A. T., O’Haver, T. C., and Keliher, P. N., Anal. Chem., 1976, 48, 1166. Fowler, W. K., Knapp, D, O., and Winefordner, J . D., Anal. Chem., 1974, 46, 601. Lipari, F., and Plankey, F, W., Anal. Chem., 1978, 50, 386. O’Haver, T. C . , Epstein, M. S., and Zander, A. T., Anal. Chem., 1977,49, 458. McWilliam, I. G., J. Sci. Instrum., 1959, 36, 51. Koirtyohann, S . R., Glass, E. D., Yates, D. A., Hinderberger, E. J., and Lichte, F. E., Anal. Chem., Michel, R. G., Hall, M. L., Ottaway, J . M., and Fell, G. S., Analyst, 1979, 104, 491. Michel, R. G., Hall, M. L., Rowland, S. A. K., Sneddon, J., Ottaway, J. M., and Fell, G. S., Analyst, Peaston, R. T., Med. Lab. Technol.. 1973, 30, 249. Roldan, R., Rev. Sci. Instrztm., 1969, 40, 1388. Ottaway, J. M., Bezur, L., and Marshall, J., Analyst, 1980, 105, 1130. Young, M., in “Optics and Lasers,” Springer-Verlag, New York, 1977. Johnson, D. J., Plankey, F. W., and Winefordner, J. D., Anal. Cham., 1975, 47, 1739. 101, 753. 1977, 49, 1121. 1979, 104, 505. Received September 17th, 1979 Accepted November 13th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9810600288

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Analyst, March, 1981, Vol. 106, pp. 299-304 299 Determination of Lead in Urine by Atomic-absorption Spectroscopy with Electrothermal Atomisation D. J. Hodges and D. Skelding Research and Development Department, The Associated Octel Company Limited, P.O. Box 17, Oil Sites Road, Ellesmere Port, South Wirral, L65 4HF A method is described for the determination of lead in urine by atomic- absorption spectroscopy with electrothermal atomisation. The determina- tion is rapid and minimum pre-treatment of the sample is required. Matrix interference is minimised by the addition of orthophosphoric acid and also by pre-coating the graphite tubes with molybdenum. Consideration is given to possible losses of organically bound lead during the drying and ashing cycle and steps to prevent these are incorporated in the procedure.The range of the method is 5-200 pg 1-1 of lead. Keywords : Lead determination ; urine ; electrothermal atomic-absorption spectroscopy ; matrix interference A method for determining lead in urine was sought that was easy to operate, adaptable to large numbers of samples and of an accuracy adequate for use as a screening test for likely exposure. The procedure used in these laboratories for many years was that described by Bambach and Burkey,l which was superseded by the semi-automatic method of Browett and Moss.2 Both of these procedures included time- and labour-intensive wet- and dry-ashing stages. There was also the problem of the large sample volume (50-100 ml) required in these methods, which with “snap” samples precluded a second aliquot being taken if necessary owing to accidental loss or contamination of the original aliquot.Procedures available for the determination of small concentrations of lead in urine include spectrophotometric, polarographic and flame atomic-absorption techniques. All methods for the determination of lead in urine by flame atomic-absorption spectroscopy are based on solvent extraction after complexing the lead with reagents containing the dithiocarbamate group.3 However, incomplete extraction of endogenous lead has been observed4 unless some method of pre-ashing the sample is undertaken. The analysis time required by these methods of treatment would not be acceptable for our purpose. The use of carbon furnace atomic-absorption spectroscopy for the determination of lead in urine offered the possibility of a minimum of sample pre-treatment provided that the following difficulties could be overcome : interference from anions and cations, non-specific absorption and loss of lead by volatilisation either as halide or as organically combined lead.Other workers have had varying degrees of success in eliminating these problems when dealing with natural water samples. Alkaline earth metal interference has been minimised by the addition of ascorbic acid.5 Matrix modification achieved by the addition of ammonium nitrate6 has been used to remove the interference of high chloride concentrations. However, none of these methods was effective when applied to urine, which exhibits a combination of these interference effects.A matrix modification technique involving the addition of orthophosphoric acid and pre- treatment of the carbon tube with ammonium molybdate has been found’ to be effective in minimising the interferences encountered when analysing urine. Experiment a1 Apparatus A Varian 175B atomic-absorption spectrometer was used together with a CRA-90 atomiser fitted with an ASD-53 automatic injection system. The diluted samples were mixed using a Denley Spiramix 5. The warming plate was constructed from an aluminium block with 48 indents to hold the cups. Electric heating elements were placed in the base of the block and the temperature was controlled using a Pye Ether Mini. The block was maintained at a300 Analyst, Val. 106 surface temperature of 40 “C. Gilson adjustable-volume and Eppendorf fixed-volume disposable-tip pipettes were used for transfer of samples and reagents.Borosilicate glass cups fitted with polythene caps were used for sample preparation. HODGES AND SKELDING: LEAD IN URINE BY AAS Reagents All reagents were of at least AnalaR grade. The orthophosphoric acid was of Aristar grade (BDH Chemicals). Ammonium molybdate solution, 5% m/V. Dissolve 5 g of ammonium molybdate in approximately 60 ml of de-ionised water, add 5 ml of orthophosphoric acid, mix, transfer into a 100-ml calibrated flask and make up to volume with de-ionised water. Dissolve 1 g of ammonium molybdate in about 60 ml of de-ionised water, add 1 ml of orthophosphoric acid, mix and make up to 100 ml in a calibrated flask. Ascorbic acid solution, 25% m/V.Dissolve 6.25 g of ascorbic acid in de-ionised water and make up to 25 ml in a calibrated flask. “Urine diluent.” To approximately 60 ml of de-ionised water contained in a 100-ml calibrated flask add 5 ml of 1% m/V ammonium molybdate solution and 2 ml of ortho- phosphoric acid. Swirl to mix, then add 1 ml of ascorbic acid solution. Dilute to volume with de-ionised water and shake to mix. The solution will turn a royal blue colour on standing. Dissolve 16.6g of potassium iodide in approximately 80ml of de-ionised water contained in a 100-ml calibrated flask; add 12.7 g of iodine and shake to dissolve. Make up to 100 ml with de-ionised water and shake well to mix. Dissolve 0.1599 g of lead(I1) nitrate in approximately 80 ml of de-ionised water; add 1 ml of distilled concentrated nitric acid. Transfer the solution quantitatively into a 100-ml calibrated flask, make up to volume with de-ionised water and shake to mix.Transfer 125 pl of the stock solution into a 25-ml calibrated flask containing approximately 20ml of de-ionised water. Make up to volume with de- ionised water and shake thoroughly to mix. This solution contains 5 pg ml-l of lead. Ammonium molybdate solution, 1% m/V. Iodine solution, 1.0 N. Standard lead solution. This stock solution contains 1 mg ml-l of lead. Dilute standard lead solution. Procedure Tube conditioning instrument with the following parameters: Fit a standard (pyrolytically coated) carbon tube to the CRASO work head. Varian 175B s$ectrometer- Set up the Wavelength . . .. .. .. 283.3nm Lead hollow-cathode lamp.. . . 5.0mA Hydrogen lamp . . . . . . Intensity to balance Mode . . .. .. .. . . Peak concentration Dry .. .. .. .. .. 100°C, 60s Ash .. .. .. . . .. 900°C, 40s Atomiser . . .. .. . . 2100 “C, hold time 0.5 s, ramp rate 50 “C s-1 Expansion . . .. .. .. Low5 CRA-90 atomiser- ASD-53 auto samfile dispenser- Cam .. . . .. .. .. 5pl No. of injections . . .. .. 5 Transfer into a clean glass cup approximately 1.5ml of 5% m/V ammonium molybdate Set the In the course of replicate injections the back- Re-set the instrument to read background- solution, place in position 24 of the carousel and initiate the programme start. instrument to read background absorbance. ground signal will stabilise at a value close to zero. corrected absorbance. Cali bration Change the following instrument parameters : CRA-90 atomiser-March , 1981 WITH ELECTROTHERMAL ATOMISATION 301 Atomiser .. .. .. . . 2100 O C , hold time 2.0 s, ramp rate 300 "C s-l ASD-53 auto sample dispenser- Into five clean glass cups pipette 0, 5, 10, 15 and 20 pl of the dilute standard lead solution, which represents additions of 0, 50, 100, 150 and 200 pg of lead per litre of urine, respectively. Into each cup pipette 20 pl of iodine solution and 500 pl of a normal low lead content urine sample. Place the cups in the indents on the warming block and allow to stand for 5-10 min. Remove the cups from the block and add 1000 pl of urine diluent to each cup. Stopper the cups with polythene overcaps and place on the Denley mixer for 10min. Remove the caps, place sequentially in the carousel of the auto sample dispenser and initiate the programme start.Carry out reagent blanks by substituting de-ionised water for the urine. Plot the readings obtained against micrograms of lead per litre added and extrapolate the line to obtain the lead content of the urine sample. Using this value, construct a calibration graph so that the line obtained passes through the origin. No. of injections . . .. . . 3 (see Note 1) Analysis of the sample Into clean glass cups pipette 20 p1 of iodine solution and 500 pl of each of the samples. Place the clearly marked cups in the indents on the warming block and allow to stand for 5-10min. Remove the cups from the block and add lOOOp.1 of diluent to each cup. Stopper the cups with the polythene overcaps and place on the Denley mixer for 10min.Remove the caps, place in the carousel of the auto sample dispenser and initiate the pro- gramme. Read off the lead content of the sample from the calibration graph. Samples more than 24 h old and samples containing precipitate must be acidified by adding 10% V/V of distilled concentrated nitric acid. Transfer 500 pl of the acidified urine into a clean glass cup and add 65 p1 of ammonia solution (sp. gr. 0.88) to neutralise the nitric acid, then add 20 pl of iodine and proceed as above. This dilution must be taken into consideration when calculating the lead content by multiplying the number of micrograms of lead per litre (read off the calibration graph) by 1.16. NOTES- blank prior to the remainder of the calibration.intervals; this is conveniently carried out after each 25 samples. temperature of a t least 800 "C is used. 1. 2. 3. Better replication can be achieved by making a large number of injections (nine) of the urine When large numbers of samples are to be analysed the calibration should be repeated a t frequent To obtain the full benefit of the orthophosphoric acid treatment it is essential that an ashing Instrumentation The earlier work7 was carried out using a carbon tube furnace contained in a totally enclosed cell with silica windows, which, over a number of determinations, tended to fog owing to the volatile products formed during the various stages of the analysis. The effect of this was not only a gradual drift in the base reading (which could be compensated for to some extent) but more seriously large and unpredictable jumps in the base-line absorbance reading arising from the thermally unstable nature of the deposit.The presence of the deposit also reduced the effectiveness of the deuterium background correction. It was therefore necessary to clean the silica windows, which involved dismantling the cell after a comparatively small number of sample analyses. As it was intended that the final instrumentation should incorporate an automatic injec- tion system, the formation of the deposit on the windows could not be tolerated. Various methods of overcoming this problem were tried without success. It was decided, therefore, to change the instrumentation to incorporate an open-type carbon furnace with a laminar-flow inert gas shield and an automatic sample injector.All of the present work was carried out using the Varian equipment. However, later versions of both Instrumentation Laboratory and Perkin-Elmer carbon furnaces incorporate modifications to reduce fogging of the side windows so that the analytical scheme is not necessarily restricted to any particular instrument.302 Analyst, VoZ. 106 Calibration Calibration graphs were constructed by adding lead nitrate to urine and to distilled water using the method given but without the addition of iodine or ascorbic acid. After sub- tracting the reagent blank and lead already present in the urine the. calibration graphs shown in Fig. 1 were obtained. These show large differences in slope and demonstrate the necessity of using a calibration of standard additions to a urine sample.In the earlier study? where lead calibration data were based on peak-area measurements, little difference in slope was observed. Because of its low surface tension the diluted urine spreads over a comparatively large area when applied to the graphite tube. Consequently, on atomisation the concentration of atoms within the light beam are more diffuse than when urine is omitted. Errors due to this effect are less when peak-area measurements are taken. With the instrument used in this work it was only practical to measure peak heights. How- ever, other factors in design (as mentioned above under Instrumentation) were more important when considering the choice of instrument. HODGES AND SKELDING: LEAD IN URINE BY AAS al a Added lead/pg I-’ Fig.1. Calibration graphs for lead in urine. A, Pba+ added to distilled water; and B, Pbz+ added to low lead content urine sample. Recovery of Organically Bound Lead The form in which lead is excreted in urine has been studied by several workers8s9 and it is likely that at least part of the lead is present as a complex with organic materials normally present as waste products. Where there has been exposure to alkyllead the lead may be excreted partly in an organic form.l0,l1 Further, treatment of lead poisoning by chemo- therapy invariably results in elevated levels of lead in urine and it is almost certain that in these instances the lead is excreted as a complex with the agent used. As the proposed method includes digestion and finally ashing with orthophosphoric acid, any organically bound lead complex in the sample will be decomposed before atomisation.However, it is possible that some lead could be lost as a relatively volatile organic compound during the heating cycle before sufficient destruction of the complex has taken place. Most organic complexes of lead can be decomposed by oxidation with potassium perman- ganate, nitric acid, bromine, iodine, etc. Of these, potassium permanganate reacts slowly and it was thought that it might have some lasting effect in the carbon tube furnace; nitric acid normally requires standing at room temperature for some time followed by evaporation to dryness; bromine is very effective, reacting quickly in the cold, but it requires strict safety measures in handling.Woessner and CholaklO used iodine to decompose organically bound lead in their rapid screening method. Because of the detrimental effects of free halogen on the carbon tube furnace, it was considered necessary to reduce any excess of iodine before injection of the samples. The choice of reagents for achieving this was fairly wide and included sodium sulphite,1° hydroxylammonium or hydrazinium chloride and ascorbic acid. Sodium sulphite was rejected because it would introduce increased salt levels into the solutions and the carbonMarch, 1981 WITH ELECTROTHERMAL ATOMISATION 303 furnace. Hydroxylammonium and hydrazinium chlorides were slow in action and also could not be incorporated easily into the diluent. Ascorbic acid, on the other hand, effects a rapid reduction of excess of iodine and the reaction products present no problems.Ascorbic acid can be incorporated in the diluent make-up. On standing, the diluent turns dark blue, presumably owing to the formation of a molybdenum oxide complex, but this does not interfere with its action. In order to assess the effectiveness of the iodine treatment in preventing loss of volatile lead complexes, it was decided to determine the recovery of inorganic lead added to urine in the presence of EDTA and D-penicillamine, which are used in the chemotherapy of lead poisoning and are possible excretion products during this treatment. Also, the recovery of lead added to urine as trimethyllead chloride was determined. In the presence of EDTA and D-penicillamine, lead nitrate was added to a low lead content urine to give lead concentrations in the range 50-200 pg 1-1 together with at least a 4-fold excess of the possible interferent.The recovery of trimethyllead was determined over the same range. The procedure was carried out first using the conditions given in the method and then omitting the iodine and ascorbic acid from the treatment. The results are shown in Table I and the general conclusion is that oxidation with iodine is necessary in order to obtain complete recovery of organically bound lead. TABLE I RECOVERY OF ORGANICALLY BOUND LEAD Mean recovery over the range 50-200 pg 1-l f standard deviation, yo Conditions $be+ + EDTA Pb2+ + D-penicillamine Me,PbCl ' Iodine treatment at 40 "C . . . . 99 f 1.6 102 If 1.1 96 f 4.7 No iodine at ambient temperature .. 96 f 4.0 103 f 2.4 15 f 2.2 The recovery of trimethyllead was lowest at 200 pg 1-1 of lead (91%), and this was further examined by lengthening the heating period and increasing the temperature of the oxidation stage in steps to 30 min at 60 "C without improving the recovery at this level. Because these lead concentrations are higher than normally experienced the recovery was deemed to be adequate. Results The proposed method was compared with that currently in use,2 which involves nitric acid digestion and furnace ashing of the sample followed by determination of the lead by dithizone automatic spectrophotometric procedure. The comparative study extended over a period of 3 months and involved 2559 samples taken in the course of routine medical examination of workers engaged in the manufacture of lead alkyls.The correlation coefficients and regression line equations are given in Table 11. TABLE I1 STATISTICAL COMPARISON OF PROPOSED ATOMIC-ABSORPTION METHOD WITH CURRENT SEMI-AUTOMATIC DITHIZONE METHOD^ No. of Correlation Month samples coefficient Regression* 1 759 0.961 3 AAS = 0.948D + 1.873 2 830 0.973 6 AAS = 0.971D + 1.274 3 970 0.986 1 AAS = 0.972D + 0.405 * AAS = result by atomic-absorption method; D = result by dithizone method. The repeatability of the method was tested by the analysis of six urine samples using ten The range of lead content was chosen to cover the normal calibra- separate aliquots of each. tion levels. The results are given in Table 111.304 HODGES AND SKELDING TABLE 111 METHOD REPEATABILITY DATA Sample No. 1 2 Mean lead contentlpg I-' .. .. . . 108.5 61.3 Standard deviationlpg 1-1 (n = 10) . . 6.29 4.03 The data given in Fig. 2 typically indicate the range method was designed. 3 4 5 6 168.1 79.9 19.2 118.8 3.38 5.80 2.04 6.07 of lead in urine levels for which the Conclusion The proposed method for the determination of lead in urine is raDid and convenient to use, wfiilst being economic as regards time and reagent costs. Onla routine basis 50-60 samples can be processed, together with calibration standards, in 4 h. The accuracy and precision of the analytical result is comparable to that obtained by more established pro- cedures. NON-EWOSED W 10 26 38 46 58 68 70 313 PO 188% , , , , . . . , , . , I 128 TO 140 N Fig. 2. Histograms showing levels of lead in urine of exposed and non-exposed workers during a period of 1 month. Acknowledgements are made to The Associated Octel Company for permission to publish this paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. References Bambach, K., and Burkey, R. E., Ind. Eng. Chem., 1942, 14, 904. Browett, E. V., and Moss, R., Analyst, 1965, 90, 715. Willis, J., Anal. Chem., 1962, 34, 614. Selander, S., and Cramer, K., BY. J. Ind. Med., 1968, 25, 139. Regan, J. G. T., and Warren, J., Analyst, 1976, 101, 220. Ediger, R. D., Peterson, G. E., and Kerber, J. D., At. Absorpt. Newsl., 1974, 13, 61. Hodges, D. J., Analyst, 1977, 102, 66. Nelson. K. W.. and Hamm. R. E., Arch. Ind. Health, 1958, 17, 38. Dinischiotu, G. T., Nestorescu, B., Radulescu, I. C., Ionescu, C., Preda, N., and Ilutza, G., BY. Woessner, W. W., and Cholak, J., Arch. Ind. Hyg. Occup. Med., 1953, 7 , 249. Bolanowska, W., Piotrowski, J., and Garczyfiski, H., Arch. Toxikol., 1967, 22, 278. J . Ind. Med., 1960, 17, 141. Received May 30th, 1980 Accepted September 19th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9810600299

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Analyst, March, 1981, Vol. 106, pp. 305-310 305 Determination of Tin in the Presence of Lead by Stripping Voltammetry with Collection at a Rotating Ring-disc Electrode P. Kiekens, H. Verplaetse, L. De Cock and E. Temmerman Laboratory for Analytical Chemistry, Ghent University, Jozef Plateaustraat, 22 9000 Ghent, Belgium Stripping voltammetry with collection at a glassy carbon rotating ring-disc electrode is used for the determination of tin in the presence of lead. Tin can be readily determined at a level of 1 p~ (0.12 pg ml-l) in the presence of a 50-fold concentration of lead. Determinations down to 0.25 p~ (0.03 pg ml-l) are possible with prolonged deposition times (about 30 min). Keywords : Stripping voltammetry ; rotating ring-disc electrode ; tin deter- mination ; interference by lead A major difficulty in determining tin by anodic-stripping voltammetry is the serious inter- ference by lead in most electrolytes.Various methods have been suggested to solve this problem. A preliminary separation can be performed by extraction,' distillation,2 or co- precipitation of lead3; also less common supporting electrolytes may be u ~ e d . ~ , ~ In this study the possibility was examined of determining tin in the presence of lead by stripping voltam- metry with collection at a rotating ring-disc electrode constructed of glassy carbon. This application was introduced by Johnson and Allen,6J who determined silver and mercury down to 10-lo and 5 x 10-8 M, respectively. Its possible use in trace metal determination (of copper, lead and zinc) has also been demonstrated by Laser and Ariel.8 The method consists of electro-deposition of the species to be determined at the disc of a rotating ring-disc electrode (RRDE) followed by stripping of the deposit using a linearly varying potential.A fraction of the oxidised species is transported to the surrounding ring electrode where a selective detection is possible at a fixed potential. In the acidic medium used it was found that the oxidation product at the disc, tin(II), can be further oxidised to tin(1V) at an appropriate positive ring potential. No reaction and thus no interference by lead(I1) occurs at this potential. As the potential of the ring is constant, almost no capacitive current is flowing and the base line is essentially flat. This results in an improved accuracy and sensitivity.6 Moreover the latter can be enhanced by the use of a high scan rate (v) at the disc, resulting in a higher dissolution current without influencing the base line at the ring.The charge equivalent with the oxidation current of the tin(I1) species collected at the ring is given by the following equation,6 which can be derived from the Levich equationg where A is the surface area of the electrode (cm2), v the kinematic viscosity (cm2 s-l), n the number of electrons involved, D the diffusion coefficient (cm2s-l), w the rotation speed (radians s-l), cb the bulk concentration of the species determined (molml-l), Tdep the accumulation time (s) and N the collection efficiency.1° The collection efficiency, N , represents the fraction of the electroactive species produced at the disc that reacts at the ring, part of the reaction product at the disc escaping into the bulk of the solution during the crossing of the gap betwc The value of N can be calculated from the dimensions of the ring-dis :trade.ing and disc. Experimental Reagents All solutions were prepared from analytical-grade reagents and from water freshly generated by a Milli-Q system (Millipore Inc.). Stock solutions of tin(I1) were prepared by dissolving tin(I1) chloride dihydrate in Experiments were performed in 1 M hydrochloric acid.306 Analyst, Val. 106 concentrated hydrochloric acid to avoid hydrolysis. Lead( 11) solutions were prepared by dissolving lead( 11) chloride in 1 M hydrochloric acid. Solutions were carefully de-aerated with nitrogen containing less than 1 p.p.m.of oxygen. KIEKENS et d. : TIN IN THE PRESENCE OF LEAD BY Instrumentation A home-made ring-disc electrode of glassy carbon (Tokai Electrode Co., Tokyo) was used, with a disc radius of 0.262 cm and inner and outer ring radii of 0.273 and 0.330 cm, respec- tively. This gives a = 0.131 and ,B = 0.867.1° The collection efficiency determined with the copper(I1) - copper(1) system in 0.5 M potassium chloride solutionll was 0.33. The electrolysis cell was a perspex vessel of about 150-ml capacity in which the glassy carbon ring-disc electrode was centrally placed. The auxiliary electrode (platinum) was placed in a compartment separated from the cell by a fritted glass disc. A saturated calomel electrode (SCE) served as reference electrode and was connected to the cell via a salt bridge and a U-tube containing 1 M hydrochloric acid and saturated potassium chloride solution, respectively.Con- tinuous selection of speeds from zero to 315 radians s-1 was possible. The rate was strictly controlled using a servo control unit (Tacussel Asservitex). The rotation rate was measured by means of a proximity probe (Philips PR 9373) and frequency meter (Philips PM 6601). Potential control was maintained by means of a Tacussel sweep generator GSATP and potentiostat Bi-pad. Curves were recorded on a Hewlett-Packard X - YY' recorder, Model HY 7046A. A water-bath allowed the temperature to be held constant at 20 "C. Electrode rotation was achieved by a motor from Brion-Leroux (Birotax, Type I).The areas of the recorded peaks were determined with a planimeter. Electrode Pre-treatment The glassy carbon ring-disc electrode was polished by standard metallographic techniques. Final polishing used 0.05 pm alumina on Buehler microcloth to a mirror-like smoothness. After being rinsed with water the polished electrode was activated in 1 M hydrochloric acid by cycling the potential between 1 .O and - 1.2 V against the SCE at a scan rate of 0.030 V s-l until a reproducible and very low background was obtained. The electrode was then held successively at -0.5 and 0.OV against the SCE for 5 min to remove any oxide film that might have formed and metallic impurities that might have been deposited at -0.5 V, respectively. Nitrogen was passed through the solution during this pre-treatment, after which the activated electrode was transferred into the already de-aerated test solution.Between successive recordings the disc electrode was cleaned by anodic polarisation at a potential of 0.6 V for 2 min. No special treatment of the ring was necessary. Analytical Procedure for Electrolysis Pre-electrolysis was started by switching the potential of the disc to the desired Edep. During deposition the potential of the ring, Er, was held at 0.0 V. One minute before the expiration of Tdep, E, was switched to the appropriate collection potential. After deposition the disc potential was scanned in the anodic direction to 0.6 V. The electrode was kept at this potential for about 2 min during which time nitrogen was passed through the solution.The electrode was then ready for another run. After collection E , was returned to 0.OV. The blank was considered to be the charge or peak current obtained for Tdep equal to zero, immediately after a scan with finite deposition time. The initial disc potential (Ei) was 0.6 V against the SCE. Results and Discussion Current - potential graphs for the oxidation of tin and lead in hydrochloric acid with the simultaneous collection of tin(I1) at the ring are presented in Fig. 1. When the current at the disc varies with time, the ring current reflects the fraction of the tin(I1) ions produced at the disc that is collected at the ring (given by the collection efficiency N). The ring to disc current ratio is no longer equal to N when lead is present and is oxidised at the disc but not collected at the ring.The shift of the ring collection peak to a more positive value can be explained by the high scan The anodic peak at the disc is seen at approximately -0.45 V.March, 1981 STRIPPING VOLTAMMETRY USING A ROTATING RING-DISC ELECTRODE 307 rate and the finite time necessary for the tin(I1) species leaving the disc to cross the gap between ring and disc.6 The unequivocally definable base line for the ring collection peak is notable. The supporting electrolyte was always 1 M because smaller peaks were obtained both at lower and higher hydrochloric acid concentrations. The latter also causes an increased screening effect at the electrode surface because of the enhanced hydrogen evolution. Fig. 2 shows a plot of Qr as a function of Edep for Er = 1 V against the SCE.Fig. 3 represents the same for various Er values at constant Edep. From the shapes of the curves in Figs. 2 and 3, it appears that the oxidation at the ring proceeds more irreversibly compared with the reduction (deposition) at the disc. A suitable collection potential at the ring and deposition potential at the disc are, respectively, +1 and -1 V against the SCE. Accumulation potentials that are too negative cause lower peaks, probably due to screening of the electrode surface by hydrogen evolution. The oxidation state of the tin species responsible for the oxidation current observed at the ring was determined with a solution containing only tin(I1) by comparing the ring currents obtained at collection potentials of +1 and -1 V against the SCE.The values of Qr for both oxidation and reduction were nearly the same, indicating that tin(I1) is mainly formed during the stripping process. Equation (1) predicts that Qr is proportional to the tin(I1) bulk concentration, to the deposition time, Tdep, and to the square root of the rotation speed of the electrode. Fig. 4 shows the linear dependence found between Q r and w1I2 for a solution 10 p~ in tin(I1) and Disk 0 -0.2 -0.4 -0 > --0.8 EdN versu: SCE Fig. 1. Current - potential urves for stripping voltam- metry v ,th collection in 1 rl hydrochloric acid at the glassy carbon electrode. Con- centration of tin(I1) = 10 p~ ; concentration of lead(I1) = 2 p ~ ; E d e p = -1 v VeYSUS SCE; E, = 1 v versus SCE; T d e p = 5 min; v =0.1 V s-l; w = 188 rad s-l.Dotted line represents constructed baseline. -0.5 -0.7 -0.9 -1.1 EdN versus SCE Fig. 2. Dependence of Qr on E d e p . Con- centration of tin(I1) = 20 p ~ ; E, = 1 V versus SCE; T d e p = 5 min; v = 0.1 V s-l; w = 157 rad s-l.308 KIEKENS et al.: TIN IN THE PRESENCE OF LEAD BY Analyst, Vol. 106 1.2 1.0 0.8 0.6 0.4 E, N versus SCE Fig. 3. Dependence of Qr stripped from the glassy carbon disc electrode on Er . Concentration of tin(I1) = 2 0 p ~ ; E d e p = -1 V Venus SCE; T d e p = 5 min; 21 = 0.1 V s-l; o = 188 rad s-l. 5 p~ in lead(I1). The dependence of Q r on Tdep is shown in Fig. 5. For curve A the solution was also made 2 p~ in lead(I1) whereas for curve B only tin(I1) was added. It is clear that the presence of lead improves tin deposition, probably by providing precipitation centres for the tin.Fig. 6 shows Q r and I r as functions of the tin(I1) concentration in solu- tion. Neither curve passes through the origin. Small deviations from linearity are also Fig. 4. Dependence of Qr on ~ 1 ' 2 . Concentration of tin(I1) = 1 0 p ~ ; con- centration of lead(I1) = 5 p ~ ; Edep = - 1 V against SCE; Er = 1 V; T d e p = 5 min; o = 0.1 V s-1. 500 400 200 100 0 10 20 30 Td,,/m i n Fig. 5. Dependence of Qr on electro- lysis time. E d e p = -1 V vwsus SCE; E, = 1 V ; v = 0.1Vs-l; and o = 188 rad s-l. A, Concentration of tin(I1) = 1 0 p ~ and concentration of lead(I1) = 2 p ~ . B, Concentration of tin(I1) = 10 pM.March, 1981 STRIPPING VOLTAMMETRY USING A ROTATING RING-DISC ELECTRODE 309 200 Y & 100 0 400 3 200 5 10 15 20 Tin (11) concentration/M x lo6 Fig.6. Dependence of Q r ( x ) and I , (0) on the concentration of tin(I1). Concentration of lead(I1) = 5pM; E d e p = - 1 v VeYSysz&S SCE; Er = 1 v . T d e p = 6 min; D = 0.1 V s-1; o = 157 rad s-1. observed at low concentrations of tin(I1) and short deposition times. Ring and disc peak areas were independent of scan rate while the collection peak heights showed a linear dependence on the scan rate, v, between 0.010 and 0.100 V s-l. Higher scan rates caused deviations, such as broadening of the peaks. As lead is often present at concentrations higher than those of tin, the influence of adding an excess of lead was also investigated. Fig. 7 shows the dependence of Qr on lead(I1) concentration at constant tin(I1) concentration. Lead does not interfere when the ratio of the lead(I1) to tin(I1) concentrations does not exceed about 50.At higher ratios results become irreproducible and also split collection peaks are observed. At Er = 1 V against the SCE, tin is determined; at Er = -1 V both lead and tin re-deposit on the ring and the concentra- tion of lead(I1) can be evaluated from the difference between the Qr values. A t tin(I1) concentrations below 1 p~ increased deposition times are necessary and down to 0.25 p~ of tin(I1) can be determined. In this concentration range determinations are less reproducible and accurate, and more time consuming. Fig. 8 shows the dependence of I , on the concentration of tin(I1) at concentrations lower than 1 PM. Satisfactory linearity is found down to tin(I1) concentrations of 2.5 x 10-7 M when long deposition times are used.I t is also possible to evaluate the amount of excess of lead. & 50 - 2 - 1 0 1 2 concentration tin II LO^ [ concentration leal (11, ] Fig. 7. Dependence of Qr of tin on the tin(I1) to lead(I1) concentration ratio for 10 p~ tin(I1). 5min; v = 0.1 Vs-1; o = 188rads-'. E d e p = -1 v ZteYSUS SCE; Er = 1 v; T d e p =310 KIEKENS, VERPLAETSE, DE COCK AND TEMMERMAN 0 5 10 Concentration of tin (it) x IO’/M Fig. 8. Dependence of Ir on bulk concentration of tin(I1). 1 V; T d e p = 30 min; v = 0.1 V s-l; w = 315 rad s-l. Edep = - 1 V versus SCE; Er = Below this concentration no tin deposition could be detected, e.g., for 0.1 p~ of tin(I1) a dissolution peak could not be obtained even after accumulation times of several hours.For tin(I1) concentrations higher than 1 p ~ , the relative standard deviation of the results was about 6%. This rather high value is, in our opinion, mainly due to the integration method using a planimeter. For experimental work it is advisable to use ring peak heights, the relative standard deviation of which never exceeded about 4%. The high relative errors encountered for tin determinations, especially at concentrations below 1 p ~ , indicate the difficulties inherent in depositions at solid electrodes. Better results may be expected using a mercury film electrode2s5J2 in combination with the ring-disc technique, which is presently under study . The authors thank the N.F.W.O. for financial support. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Analytical Methods Committee, Analyst, 1967, 92, 320. Florence, T. M., and Farrar, Y. J., J . Electroanal. Chem., 1974, 51, 191. Portretnyi, V. P., Malyuta, V. F., and Chuiko. V. T.. Zh. Anal. Khim., 1973, 28, 1337. Brainina, K. Z., Talanta, 1971, 18, 513. Glodowski, S . , and Kublik, Z., Anal. Chim. Acta, 1979, 104, 55. Johnson, D. C., and Allen, R. E., Talanta, 1973, 20, 305. Allen, R. E., and Johnson, D. C., Talanta, 1973, 20, 799. Laser, D., and Ariel, M., J . Electroanal. Chem., 1974, 49, 123. Levich, V. G., “Physicochemical Hydrodynamics.” Prentice Hall, Englewood Cliffs, N. J ., 1962, Albery, W. J., and Bruckenstein, S., Trans. Faraday Soc., 1966, 62, 1920. Napp, D. T., Johnson, D. C., and Bruckenstein, S., Anal. Chem.. 1967, 39, 481. Florence, T. M., J . Electroanal. Chem., 1970, 27, 273. p. 69. Received July 14th, 1980 Accepted September 15th, 1980

ISSN:0003-2654    

DOI:10.1039/AN9810600305

出版商:RSC    

年代:1981

数据来源: RSC

摘要:

Analyst, March, 1981, Vol. 106, pp. 311-315 Direct Spectrophotometric Method for the Determination of Hydrochloric Acid-releasable 31 1 Arsenic in Sediments and Soils Shingara S. Sandhu ClafEin College, Orangeburg, South Carolina 291 15, USA A spectrophotometric method for the direct determination of hydrochloric acid-releasable inorganic arsenic has been developed and applied successfully to the quantitative evaluation of arsenic in soil and sediment samples. The method provides reliable data on the quantitative recovery of 2.0 pg of arsenic(V) added to 5 . 0 g (0.4mg kg-') of soil, clay, sand and sediment samples. The method is simple, reliable and relatively rapid; 24 samples can be analysed in about 1 h. It does not require elaborate equipment and can be routinely used for the quantitative determination of arsenic in soil and soil-like material. The detection limit has been established as 0.5 pg of arsenic.The extent of ionic interference in the use of this method for arsenic determination in soil has also been quantitatively evaluated. Keywords : Arsenic determination ; soils ; sediments ; spectrophotometry Arsenic occurs naturally in the earth's crust, but a considerable amount of arsenic is added to the human environment through its uses in wood preservatives, sheep dips, fly paper, arsenical soaps, rat poison, glass additives, dye pigment for calico prints, wall paper, lead shot and pesticidexl During 1971, the estimated production of organoarsenical herbicides such as monosodium methanearsenate, disodium methanearsenate and hydroxydimethylarsine oxide (cacodylic acid) in the USA was 10.7 x 108 kg2 Generally, soils contain about 5.0 p.p.m.of arsenic, but soils with a known history of arsenic application average about 165 ~ . p . m . ~ In some places such as Buns, Switzerland, and Wiatapu Valley, New Zealand, the arsenic level in the soil may reach 104 p.p.m.4; a substantial portion of arsenic in soil and soil-like material (sediment, clay, sand, etc.) is expected to be found in soluble form and probably can be dis- lodged easily by the action of water moving through the soil. Soluble forms of arsenic are relatively more mobile in the environment596 and pose a greater potential for contaminating both ground water and surface water. Soluble forms of arsenic from soil and soil-like material are likely to enter a bioconversion chain through their initial uptake by vegetation. In order to understand the pollution potential of arsenic, it is desirable that a simple method should be made available for the quantitative evaluation of easily soluble forms of arsenic in soil and soil-like material.Various methods exist for the determination of arsenic in solutions, including spectrophoto- metr~,'-~ gas - liquid chr~matography,lO-~~ ion-exchange chr~matography,~~ coulometry,ls-le activation analysis,l9 atomic-absorption spectrophotometry,20-22 p~larography~~ and distilla- tion.24*25 The spectrophotometric method of Small and M~Cant,~ as described by Woolson et a1.,26 is widely used for the determination of total arsenic in soils. This method is tedious and time consuming because it requires 2 h after the soil sample has been digested.This method does not appear to be very precise at low levels of arsenic in soil, because the relative standard deviation for the recovery of 7.5 pg of arsenic added to 1 g of soil (7.5 p.p.m. of arsenic) is approximately 6.4%. The arsenomolybdate spectrophotometric method has never been applied to the determination of easily soluble arsenic in soil prior to its digestion. This paper presents a simple, direct and rapid method for the quantitative determination of easily soluble forms of arsenic in soil, clay, sand and sediment samples at parts per billion (lo9) levels. This technique differs from the spectrophotometric method of Small and M ~ C a n t ~ . ~ ~ in that the colour for photometric measurement is developed directly by passing the arsine from the generator through the silver diethyldithiocarbamate reagent.This eliminates three tedious steps involved in the previous m e t h ~ d ~ s ~ ~ : (a) absorbing the arsine in an ice-cold iodine trap, ( b ) treating it with colour-developing reagents and (c) heating the mixture for 1 h for colour development prior to photometric measurements. This results in a considerable saving of time.312 SANDHU : SPECTROPHOTOMETRIC DETERMINATION OF Analyst, VoZ. 106 Experimental Apparatus previou~ly.~~ The arsine generator, absorber assembly and spectrophotometer have been described Samples The soil samples were collected from the locally exposed soil profiles of Dothan and Golds- boro series and were dried and ground to pass through a 2.0-mm sieve.Bentonite (Clay Spur, Wyoming), kaolinite (Silver Peak, Nevada), and sand (Ottawa flint, silica sand) samples were procured from Wards Natural Science Establishment, Rochester, N.Y. Sediment samples were collected from the Edisto River and the Caw Caw Swamp, Orangeburg, South Carolina. These samples were also dried and ground to pass through a 2.0-mm sieve. Solutions Stock solutions of arsenic containing an ionic concentration of 1 g 1-1 were prepared from analytical-reagent grade sodium arsenate (Na2HAs0,.7H,0) and sodium cacodylate [(CH,)asO,Na]. Intermediate solutions were prepared by diluting the stock solutions 1 + 10. Working solutions containing requisite concentrations of arsenic were obtained by diluting the intermediate soIut ions. Procedure The native forms of hydrochloric acid-releasable arsenic in soil, clay, sand and sediment were determined in the untreated samples.Ten grams of each of these materials, followed by 50.0 ml of de-ionised water, 7.5 ml of concentrated (12 M) hydrochloric acid (the volume of acid used varied slightly with the type of soil analysed), 2.0 ml of potassium iodide solution (150 g 1-1) and 0.7 ml of tin(I1) chloride solution (400 g 1-1 of SnC&.2H20 in concentrated hydrochloric acid) were transferred into an arsine generator and allowed to stand for about 30 min, with occasional manual shaking, in order to reduce the arsenic to the trivalent state (the hydrochloric acid concentration of the mixture in the generator is about 1.50 M).Three grams of 20-30-mesh zinc were added to the arsine generator, and the generator was immediately connected to the absorber assembly, which was equipped with a lead acetate-impregnated glass-wool scrubber and contained 4.0 ml of silver diethyldithiocarbamate (SDDC) reagent [I g of AgSCSN(C,H,), in 200 ml of pyridine]. The arsine reacts with the SDDC and produces a red complex, which is measured at 535 nm against a reagent blank (SDDC solution treated in the absorber tube according to the experimental procedure, but without soil or arsenic). An absorbance calibration graph using 0.0, 1.0, 2.0, 4.0 and 5.0 pg of arsenic was prepared. A 100-g amount of each of the soil, clay, sand and sediment samples (triplicate determina- tions were performed) was transferred into 250-ml Pyrex glass evaporating dishes. A 4-ml volume of a 10.0 mg 1-1 arsenic solution prepared from sodium arsenate and sodium cacodylate was added, a drop at a time, to each sample, the sample being constantly stirred with a plastic spatula during the addition of solution. The stirring was continued until a homogeneous mixture was obtained.The samples, treated with arsenic solution, were spread on clean filter-papers, which were placed on the laboratory bench for uniform drying overnight. The following day (about 24 h later), the samples were transferred into the original evaporating dishes and ground with small pestles. Every possible effort was made to achieve a homo- geneous mixture. The samples were transferred into 4-02 bottles for storage and future use.Five grams of each of these samples (triplicate determinations were performed) were transferred to the arsine generator for the determination of total arsenic in accordance with the procedure described previously. The arsenic recovery was obtained by subtracting the amount of native arsenic from the total amount. Two hundred grams of Goldsboro,, soil was also treated with arsenic(V), chromium(V1) (K,Cr,O,), copper(I1) [Cu(N0,),.3H20], nickel(I1) (NiC12.6H,0), antimony(II1) (SbC1,) and mercury(I1) (HgC1,) solutions so that 5.0 g of soil contained 6.0 pg of arsenic(V) and various concentrations of the individual metal ions as interferents.March, 1981 HYDROCHLORIC ACID-RELEASABLE ARSENIC Results and Discussion 313 The method described here is a modification of the standard method% recommended for the determination of arsenic in water and waste water.The detection limit for water samples has been established as 0.5 pg of arsenic.27 Only the inorganic arsenic in water samples is reduced to arsine (ASH,) during the zinc - acid reaction.% Organically bonded arsenic is expected not to be released unless the water samples are oxidised. Consequently, the spectrophotometric method reported here does not determine the total arsenic in soil, clay, sand and sediment samples ; instead, it provides information on hydrochloric acid-releasable inorganic arsenic. TABLE I NATIVE ARSENIC AND ARSENIC RECOVERED FROM SOIL, CLAY, SAND AND SEDIMENT SAMPLES Sample DothanAp . . .. .. Dothan*, . . .. * . DothanB,, . . .. ..do than^,, . . .. .. GoldsboroAp . . .. .. GoldsboroB,, . . .. .. Goldsboros,, . . .. .. GoldsboroB,, . . .. .. Bentonite . . .. .. Kaolinite . . .. .. Sand .. .. Sediment (Ndrih Edisto River) Sediment (South Edisto River) Sediment (Caw Caw Swamp) .. .. .. .. .. .. .. .. .. .. .. .. .. .. Arsenic concentration/pg per 5 g I A Recovery, Native* Added Total* Recoveredt yo 1.09 2.0 3.10 2.01 100.00 0.28 2.0 2.25 1.97 98.5 0.36 2.0 2.33 1.97 98.5 1.50 2.0 3.45 1.95 97.5 0.38 2.0 2.33 1.95 97.5 0.50 2.0 2.40 1.90 95.0 0.59 2.0 2.63 2.04 102.0 0.98 2.0 2.90 1.92 96.0 6.60 2.0 8.60 2.00 100.0 0.92 2.0 3.00 2.08 104.0 DS 2.0 1.96 1.96 98.0 1.02 2.0 3.05 2.03 101.5 1.27 2.0 3.18 1.91 95.5 3.16 2.0 5.13 1.97 98.5 Standard deviation, "/o 4.6 5.3 4.5 4.8 4.6 5.0 4.7 4.2 4.7 4.4 4.3 4.9 4.2 4.3 * Mean of ten determinations.t Recovered = Total - Native. Detected but not quantified. The amounts of native arsenic found in soil, clay, sand and sediment samples are given in Table I. I t appears that the proposed spectrophotometric method is able to detect and quantify trace amounts of arsenic contaminations in soil and soil-like materials (0.056 pg gl). Bentonite appears to be relatively high in hydrochloric acid-releasable arsenic, whereas the sand samples show very small, non-quantifiable arsenic contamination. The evaluation of acid-releasable arsenic in sediment also appears to be quantitative. The amount of arsenic(V) added to an individual sample and the total arsenic determined by the proposed method are also given in Table I. The amount of arsenic recovered was determ- ined by subtracting the amount of native arsenic from the total hydrochloric acid-releasable arsenic in the sample.All calculations were based on 5.0 g of sample used for assay. The recovery of arsenic added to soil, clay, sand and sediment samples varied from 104% for kaolinite to 95% for Goldsboro,,,. It appears that the proposed spectrophotometric method provides reliable information on the recovery of inorganic arsenic(V) added to these samples. The recovery of 10.0 pg of arsenic added as cacodylic acid to 50.0 ml of de-ionised water (0.2 mg 1-I) was 2.6% (Table 11). There appeared to be no significant change in the recovery of arsenic when the same sample was analysed 1 week later. There was no discernible recovery of arsenic achieved from 5.0 or 2.0p.g of organoarsenical that was added to 50.0ml of de- ionised water.This suggests that, although methylarsines may have been generated by cacodylic acid in de-ionised water during the zinc - acid reaction in the generator,% they do not appear to produce, under the conditions used, a red complex photometrically measurable at 535 nm. A small amount of arsenic that was recovered from 10.0 pg of organoarsenical added to 50.0 ml of de-ionised water possibly was present as inorganic arsenic in the solution prepared from cacodylic acid. The recovery of arsenic added to various soils was negligible when the samples were analysed, after treatment for 24 h with cacodylic acid. However, the recovery of arsenic from the same314 SANDHU : SPECTROPHOTOMETRIC DETERMINATION OF Analyst, VoZ.106 TABLE I1 RECOVERY OF CACODYLIC ACID ARSENIC FROM WATER AND SOIL SAMPLES Arsenic recovered* 7 Arsenic Sample added pg Water . . .. . . 10pgper 5Oml 0.26 Dothan,, .. .. 2pgper 5 g 0.03 Dothan*, .. .. 2 p g p e r 5 g 0.08 DothanB,, . . .. 2 p g p e r 5 g 0.05 Goldsboroap . . .. 2pgper 6 g 0.10 Goldsbor-,, . . .. 2pgper 5 g 0.13 24 h -7 Standard yo deviation, % pg 2.60 2.7 0.28 1.50 3.0 1.96 4.00 2.9 0.64 2.50 3.4 0.23 5.00 2.6 0.59 6.50 2.9 0.62 1 week w % deviation, % 2.8 2.4 98.0 2.8 32.0 2.6 11.5 3.2 29.6 2.8 26.0 3.1 Standard * Recovered = Total - Native (Native from Table I). set of samples improved significantly when they were analysed 1 week later. The recovery of arsenic from the soil samples is highly variable (Table 11), and appears to depend on the soil characteristics.If methylarsines do not react with SDDC, as seems apparent from the behaviour of cacodylic acid in de-ionised water, then the organoarsenical that was added to the soil must be mineralised for the generation of SDDC-intractable arsine. The recovery of organoarsenical added to various soils varied from 98.0% (in Dothan,,) to 11.5% (Dothan,,) and suggests that some soils are more effective than others in their potential to minerahse cacodylic acid. It must be mentioned that some of the inorganic arsenicals (arsenic sulphide, aluminium arsenate, etc.) are either insoluble or only sparingly soluble in 1.50 M hydrochloric acid.30 Consequently, the technique described here determines various forms of inorganic arsenic released by 1.50 M hydrochloric acid in soil-like materials.Although the interference of certain metal ions in the recovery of arsenic added to demineral- ised water has been reported previou~ly,~~ a further study on the effect of interferences on the recovery of arsenic from the soil system was conducted. No significant change in the recovery of 6 pg of arsenic(V) added to 5.0 g of soil was observed when the soil samples were also spiked with either 200 pg (40 mg k g l ) of chromium(VI), 300 pg (60 mg lcg-1) of copper(II), 300 pg (60 mg k g l ) of nickel(II), 15 pg (3 mg k g l ) of mercury(I1) or 2 pg (0.4 mg k g l ) of antimony- (111). Interference by combinations of various metal ions was also studied. It appears that the recovery of arsenic generally is not affected up to a combined metal ion concentration of 300 pg per 5 g of soil that had also been spiked with 6.0 pg of arsenic(V).Whereas higher concentrations of antimony(II1) and mercury(I1) enhance the apparent recovery of arsenic, the other metal ions at levels greater than the concentrations listed above decrease the recovery of arsenic. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. References Sub-committee on Air and Water Pollution of the Committee on Public Works, “Water Pollution,” Part 4, United States Senate, Ninety-first Congress, Second Session, US Government Printing Office, Washington, D.C.. 1970, pp. 1391-1398. Lawless, E. W., Von Rumber, R., and Ferguson, R. L., Technical Studies Department, TS00-72-04, Environmental Protection Agency, Washington, D.C., 1972.Woolson, E. A., Axley, J. J., and Kearney, P. C., Soil Sci. SOC. Am. Proc., 1971, 35, 101. Schroeder, H. A., and Balassa, J. J., J. Chronic Dis., 1966, 19, 1. Seiling, D. H., Soil Sci. SOC. Am. Proc., 1946, 11, 161. McBride, B. C., and Wolfe, R. S., Biochemistry, 1971, 10, 4312. Chaney, A. L., and Matnuson, H. J., Ind. Eng. Chem., 1940, 12, 691. Saito, J., Noyaku Seisan Gijustsu, 1962, 7 , 8. Small, H. G., Jr., and McCant. C. B., Soil Sci. SOC. Am. Proc., 1961, 25, 246. Bassett, J., Analyst. 1963, 88, 238. Ives, N. F., and Giuffrida. L., J. Assoc. Off. Anal. Chem., 1967, 50, 1. Juvet, R. S., and Wachi, F. M., A n d . Chem., 1960, 32, 290. Tadmor, J., J. Gus Chromatogr., 1964, 2, 385. Talmi, Y., and Norvell, V. E., Anal.Chem., 1975, 47, 1510. Bagliano, G., Grassini, G., and Ossicini, L., J . Chvomatogr., 1964, 14, 238. Bruckenstein, S., and Johnson, D. C., Anal. Chem., 1964, 36, 2186. Simon, K. K., Christian, G. D., and Purdy, W. C., Am. J. Clan. Pathol., 1968, 49, 207.March, 1981 HYDROCHLORIC ACID-RELEASABLE ARSENIC 315 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. Jennings, V. J., Dodson, A., and Harrison, A., Analyst. 1974, 99, 145. Lunde, G., Int. Rev. Gesamten Hydrobiol., 1967, 52, 265. Caldwell, J. C., Lishka, R., and McFarren, E., J. Am. Water Works Assoc., 1973, 56, 73. Tam, K. C., Environ. Sci. Technol., 1974, 8, 73. Fishman, M., and Spencer, R., Anal. Chem., 1977, 49, 1599. Reinke, J . , Uthe, J . F., Freeman, H. C., and Johnston, J . R., Environ. Lett., 1975, 8, 371. Ballinger, D. G., Lishka, R. J., and Gales, M. E., Jr., J. Am. Water Works Assoc., 1962, 54, 1424. Farkas, E. J., Griesbach, R. C., Schachter, D., and Hutton, M., Environ. Sci. Technol., 1972, 6, Woolson, E. A., Axley, J. J., and Kearney, P. C., Soil Sci.. 1971, 111, 158. Sandhu, S. S., and Nelson, P., Anal. Chem., 1978, 50, 322. American Public Health Association, American Water Works Association and Water Pollution Control Federation, “Standard Methods for the Analysis of Water and Waste Water,” Thirteenth Edition, American Public Health Association, Washington, D.C., 1971, p. 62. Braman, R. S., Johnson, D. L., Foreback, C. C., Ammons, J. M., and Bricker, J. L., Anal. Chem., 1977, 49, 621. Garrett, A. B., Sisler, H. H., Bonk, J., and Stoufer, R. C., “Semimicro Qualitative Analysis,” Third Edition, Blaisdell Publishing Co., Waltham, Mass., 1966. Received May 12th, 1980 Accepted September 17th, 1980 1116.

ISSN:0003-2654    

DOI:10.1039/AN9810600311

出版商:RSC    

年代:1981

数据来源: RSC