analysis — apparatus analysis and its tools
Just as the boundaries between chemical and instrumental analysis are not sharp, those between instrumental and automatic analysis are also not well defined. We can call all those processes automatic analyses in which necessary analytical procedures are carried out by machines and/or instruments without h u m a n intervention other than supervision. The results or data necessary for calculations are often drawn or pointed out by the instrument. The analytical balance is usually the first instrument one uses in an analytical procedure. Often it is used manually, although nowadays there are automatic balances available, which weigh a solid sample poured from a feeding apparatus onto the plate of the balance, and after weighing the weight is recorded and passed on to the appropriate calculator or computer. Apparatus suitable for volume measurement, dispensing and/or proportionating by volume may also be automated. A p H meter with electrodes is not an automatic device as such but when equipped with an automatic burette, switch and recorder it becomes automatic. In organic analysis we can call PregPs method for the determination of carbon and hydrogen a typical chemical procedure, because every step of the process starting with weighing of the sample up to weighing of the absorption tubes, has to be carried out by the analyst himself. In the meantime, he has to follow closely the pyrolysis of his sample, which, in general, requires both professional knowledge, experience and manual skill. Similarly, classical methods are the determinations of active hydrogen or the methoxy group. These methods, which had to be learned step by step by patient work, were used almost exclusively up to about 1960, first in macro- later in semimicroand finally in micro-scale procedures. The results were not spectacular and this type of work was not thought highly of, so that qualified chemists left this field, having been replaced by 34
laboratory assistants who, though manually skilled, did not possess the necessary background in organic chemistry, and thus their results were not too reliable. To overcome these problems, and to cope with the everincreasing number of samples, new developments had to be achieved. The first was the replacement of tube fillings in carbon-hydrogen analysis [the introduction of silver permanganate and cobalt(II,III) oxide instead of lead chromate-copper oxide etc.]. The lead(IV) oxide, used for the decomposition of nitrogen oxides, were substituted by substances such as chromic-sulphuric acid or manganese dioxide, which act even in the cold. Also, equipment has been made safer and easier to handle by using glassware with glass tops and standard joints, or by replacing glassware with stainless steel. Little happened in automation until the early 1960s. We may call "automatic" the apparatus developed during the Second World War in Switzerland by Hosli for the carbon, hydrogen and nitrogen determination. This was different from the original Pregl apparatus, in that the combustion tube and the absorbing tube, containing lead(IV) oxide were heated with well regulated electric furnace. The electrically heated, openable and moveable furnace was slowly moved by an electric motor along the combustion tube, ajid the rate of this movement had a special programme. The use of these instruments made it easier to carry out such analyses, especially in routine analyses of almost identical samples, as the pyrolysis could be carried out without supervision. Most compounds, however, had to be pyrolysed manually, and the preparation of absorption tubes and their weighing were long and delicate manual procedures. With these innovations the average time of a carbon-hydrogen determination decreased from 60 to about 30-35 min. Later were introduced automatic devices (such as the Coleman C, H and N Analyzer) which contained in a portable box all the parts of a micro apparatus suitable for the determination of carbon, hydrogen and nitrogen (according to Dumas). The combustion of the organic compound in the horizontal combustion tube was almost automatic once the weighed sample had been introduced manually. The gaseous products were passed into the absorber tubes, which had been weighed earlier and contained the same absorbent compounds as the Pregl or other types of non-automatic instruments. Thus, the instrument only partly automated the analysis. The Coleman nitrogen analyzer differed from the classical D u m a s apparatus only by the arrangement that the volume of nitrogen was not measured in an azotometer but was conducted to a piston cylinder, connected to a counter device, and so the volume of nitrogen gas was obtained automatically. The realization of automatic combustion became possible because the more effective oxidative filling materials or the high temperature empty tube 500
made possible the faster and less uniformly programmed burning of the organic compound. Hence it became unnecessary to transfer the pyrolysis products of organic c o m p o u n d s slowly, carefully, always watching the pressure of gas and the rate of gas flow, into the oxidation tube, as was necessary with the low-performance universal tube filling of Pregl. These new filler materials, as well as the empty tube, stand u p well against unequal loading, and enable one, for example, to d r o p the sample, placed in a capsule, directly into the high-temperature zone, because the conditions of the pyrolysis make incomplete pyrolysis impossible. Earlier investigations established the minimum necessary temperature and the minimum time for oxidizing quantitatively the most resistant pyrolysis products such as methane. Thus, by this means the process of pyrolysis is easy to automate, both with a continuous rate of movement or with programming, either by heating the furnace according to a suitable programme, or by dropping the organic sample into the combustion tube. In some instances the sample is weighed into a small metal container, made of an alloy which melts at the temperature of pyrolysis. The pyrolysis in some devices is carried out with a highfrequency induction furnace. It has the advantage that the wall of the combustion tube is less attacked at the high temperature used. The disadvantage, however, is that it is difficult to regulate the heating process. A more difficult problem, which was solved much later, was the automation of the quantitative determination of the pyrolysis products, such as carbon dioxide, water, nitrogen and carbon monoxide (or carbon dioxide, obtained by further oxidation). Trutnovsky [ 1 ] attempted to automate gravimetric analysis in the following way. The composition products which left the combustion tube were conducted through a thin plastic tube to the absorption tubes hanging on a microbalance. He measured the increase in weight of the tubes without manual operations. Using this method, he could carry out several carbon and hydrogen determinations in series. The method, however, was not generally adopted because of its technical difficulties. A better means of automation was to absorb the pyrolysis products in a suitable absorption liquid and to monitor some characteristic of the latter electrochemically. Some of these methods have already been dealt with in Chapter III. To obtain a more complete picture, we shall review all the important papers, notwithstanding repetition, which have contributed to the development of automatic instruments for carbon, hydrogen, nitrogen and oxygen analysis. The first electrometric method, which is still used today, was conductimetry. Korshun  described the following procedure. The organic 34*
compound is pyrolysed in a stream of oxygen using platinum as catalyst. The water is then frozen out from the gas stream, and the carbon dioxide is absorbed in 0.01 N sodium hydroxide or barium hydroxide solution. The change in conductivity of the absorber solution is monitored. After this, the frozen water is vaporized and conducted through a carbon layer, containing platinum and kept at 1000°C. The resulting gas, which now contains carbon monoxide, is oxidized to carbon dioxide on a heated copper(II) oxide layer. The carbon dioxide is then determined by conductimetry. F r o m the result the hydrogen content can be calculated. Robertson et al. [ 3 , 4 ] , Gel'man et al. , Nail and Schooley [ 6 ] , Vecera et al.  and Stuck  described modifications. Malissa and co-workers [ 9 , 1 1 ] and Stuck  dealt with the possibility of the conductimetric determination of carbon, hydrogen, sulphur and oxygen. The principle of Malissa et al.'s method is as follows. The organic compound is burned at 1200°C in a ceramic tube, which is open at one end. The carbon dioxide and water vapour are first homogenized by means of a magnetic stirrer, at constant volume, pressure and temperature. The gas mixture is separated into three equal parts with synchronized, constant-volume pumps. The first part is led into a conductivity cell, which contains 0.02 N sodium hydroxide solution, while a parallel (blank) cell contains the same reagent through which an equal volume of oxygen to that used in the combustion is bubbled. The second part of the gas is led through a reaction tube which contains calcium carbide. The acetylene formed from the water content of the gas is then burned to carbon dioxide on a copper(II) oxide layer at 550°C and the amount produced is determined in a conductivity cell. The difference between the first and second a m o u n t s of carbon dioxide is proportional to the hydrogen content of the gas. The electrical signals are converted directly into carbon and hydrogen values by means of a microprocessor, which has to be calibrated with standard samples. The developed apparatus is also suitable for the determination of sulphur . F o r this the third part of the gas can be used, which is passed through another conductivity cell containing 0.002 N sulphuric acid and hydrogen peroxide. The conductivity change of this cell gives sulphur content of the compound provided that the instrument is calibrated properly with a standard. The combustion method of Greenfield and Smith [13-15] is similar to Pregi's technique. The combustion of the organic compound is carried out in an oxygen stream using cobalt(II,III) oxide catalyst, while for the decomposition of nitrogen oxides, lead dioxide is applied. The gas flow is led into a conductivity cell containing 99.83% sulphuric acid. The conductivity of such concentrated sulphuric acid changes by the effect of the smallest amount of water. Thus, the change in the electrical conductivity is proportional to the water content of the gas and indirectly to 502
the hydrogen content of the organic compound. The gas, on leaving the cell, is conducted into another conductivity cell containing 0.05 N potassium hydroxide solution. The change in this conductivity, which is proportional to the carbon dioxide content of the gas, is recorded and is proportional to the carbon content of the sample. Gouverneur et al.  described a manometric method for the automatic determination of carbon and hydrogen. The combustion of the organic compound is carried out in oxygen with automatic control. They froze the water and the carbon dioxide separately from the gas flow with liquid nitrogen and the products obtained were led into a manometric instrument, in which the volumes of gases were measured using a Metrohm burette, which contained mercury as a sealing liquid. The method is suitable for the determination of 5-50 fig (that is, ultramicro amounts) of carbon and hydrogen . The Keidel cell was the first with which coulometry was used for the determination of carbon and hydrogen. The important part of hygrometer devised by Keidel  are two platinum wire tubes in concentric arrangement, the turns of which d o not touch. The tubes are coated with phosphorus pentoxide, protected from the outside by a plastic cover. If a 5 0 60 V D C . potential difference is applied to the tubes, and through the tubes flows a gas which contains water vapour, the phosphorus pentoxide retains the water quantitatively, which at the same time undergoes electrolysis. The quantity of electricity needed for this is proportional to the water content of the gas. At first the Keidel cell was used for the continuous determination of water content of flowing gases. Using this method we can determine 10-1000 ppm of water. As the original Keidel cell is suitable only for investigation of gases with a constant flow-rate and almost constant water content, before it was applied to the automatic analysis of organic compounds an integrator suitable for summing the electrical signal over a period of time had to be connected to it. Salzer [19, 20] determined the water content of the purified gas from the combustion tube in the Keidel cell, and the carbon dioxide content by conductimetry. The time required for one determination is 10 min and the standard deviations for carbon and hydrogen are 0.07% and 0.04%, respectively. Haber and co-workers [21, 22] modified this method, so that carbon dioxide can also be determined with the Keidel cell. First the water content is determined with the cell, then the gas is allowed to react with lithium hydroxide: 2 LiOH + C 0 = L i C 0 + H 0 2
and the gas is again led through the cell. The stoichiometry of the reaction is unfavourable as 12 parts (by weight) of carbon are equivalent to 2 parts of hydrogen. Therefore, higher sample weights and a cell of larger capacity were 503
used. The standard deviations were 0.3% for carbon and 0.03% for hydrogen. The described chemical methods for the determination of carbon and hydrogen (and sometimes oxygen and sulphur) in organic compounds were useful and necessary for the development of these automatic techniques, but they did not come into general use. Firstly, they were complicated and constant supervision was necessary, and secondly, they were later superseded by simple gas chromatography methods. These gas chromatographic methods have the disadvantage, however, that they need accurate standardization with organic compounds of known composition, and thus there is no absolute relation between the measured signal and composition, as in the case of the classical methods. Francis  and Schoniger  proposed the use of gas chromatography as a finish to elemental analysis. Later, when gas chromatographic detectors became more advanced, it was logical to use them not only for the determination of carbon dioxide and water vapour (or another gas, such as acetylene, prepared from it), but also for other gases such as nitrogen. For this, however, a different carrier gas (e.g., helium), the thermal conductivity of which is different from that of nitrogen, became necessary. The same gas is also necessary in the determination of carbon monoxide, through which the oxygen content of samples is determined. Duswall and Brandt  and Sundberg and Maresh  described the first devices based on the measurement of the thermal conductivity of gases. The sample is combusted in oxygen, on an oxidizing tube filling. The water formed is then reacted with calcium carbide to give acetylene. Together with the carbon dioxide it is first frozen out, later evaporated and, using helium as carrier gas, is introduced into a chromatographic column filled with silica gel, on which the acetylene and carbon dioxide are separated and determined. The time needed for the whole procedure is 20 min and a 10 min pause is necessary between two determinations. The apparatus can be calibrated with standard compounds and the calculation of the results is carried out from the chromatogram on the basis of peak heights or peak areas. Vogel and Quattrone  combusted the sample with oxygen in a metal b o m b inside an electrically heated platinum coil. The gas inside the b o m b was homogenized first with magnetic stirring, then the gas mixture was led with oxygen carrier gas on to a gas chromatographic column after the removal of nitrogen oxides and sulphur dioxide. The column contained diatomaceous earth moistened with dodecyl phthalate and kept at 140°C. O n this column the water was separated from carbon dioxide and their amounts were determined with a thermal conductivity detector. Walish  published the first automatic method suitable for the determination of nitrogen in addition to carbon and hydrogen. This instrument served as the basis of one type of mass-produced carbon-hydrogen-nitrogen analyzer. Walish combusted the sample in an 504
oxygen-helium mixture, fed into the apparatus at constant flow-rate. The excess of oxygen was bound on copper metal, which also decomposed the nitrogen oxides. The water vapour from the gas was bound on a heatable silica gel column, and the carbon dioxide and the nitrogen were led into one of two symmetrical katharometer cells, while the other one contained pure helium. The current of the bridge was integrated and the total carbon and nitrogen content was calculated from the signal. The gas then passed through an absorbent containing soda-asbestos followed by a second katharometer cell which gave the amount of nitrogen present. Then the silica gel column was heated and eluted with helium. A third katharometer cell then measured the amount of water and so the amount of hydrogen could be calculated. The disadvantage of the method is that the sample must not be larger than 0.5-1 mg which necessitates very careful homogenization of the sample. Sommer et al.  described a similar static integration apparatus for the determination of carbon, hydrogen and nitrogen. The combustion products were collected in an expansion chamber, the excess of oxygen was absorbed and the gas mixture was led into an evacuated system where carbon dioxide and water were selectively absorbed and finally measured in a thermal conductivity cell. The apparatus had to be calibrated with standard compounds. The method of Miller and Winefordner  is very rapid, the time required for the determination of carbon and hydrogen being 90 sec, and a similar length of time being needed for the determination of nitrogen. The sample is weighed into a small closed metal cup, which is dropped into a vertical combustion tube containing copper oxide at about 1000°C, through which oxygen flows. Water is removed by freezing out, and the gas, containing carbon dioxide, is pressed into a thermal conductivity cell where it is determined. After this the water is evaporated in a small moveable furnace and the gas, containing the water vapour, is conducted through a reaction column containing calcium hydride. From this hydrogen gas evolves which is equivalent to the amount of water present. Using oxygen as the carrier gas, hydrogen can be measured very sensitively with a thermal conductivity detector. The instrument can be calibrated with standard samples. To determine the nitrogen content the combustion of the sample is carried out in a similar way in helium containing oxygen and, after removing the excess of oxygen, the water vapour can be removed with anhydrous magnesium perchlorate, carbon dioxide with soda-asbestos, and finally the a m o u n t of nitrogen present can be determined with a thermal conductivity detector. The variation described by Pella and Colombo [32, 33] served as the basis of the Carlo Erba instrument. The sample, weighed in an aluminium vessel, was placed in the hottest part of the combustion tube, while oxygen was added to the carrier gas. The catalyst was granulated chromium(III) oxide. 505
The excess of oxygen was removed from the gas with silver-plated copper and other interfering gases were trapped with cobalt(II, III) oxide which contained some silver. Other chemisorption effects helped the purification of the gas. A thermal conductivity cell was used. F o r the determination of oxygen Gorbach and Ehrenberger  described an automatic titration apparatus, into which the carbon dioxide produced in the Unterzaucher apparatus was led. The instrument gave as accurate and precise results for the determination of oxygen as one can obtain with the gravimetric procedure, but the well-known errors of the Unterzaucher method were not eliminated. Gel'man et al.  oxidized the carbon monoxide that left the platinized carbon layer of 900°C to carbon dioxide at 300°C on a copper oxide filling. The gas stream containing the carbon dioxide was then conducted into 0.01 N sodium hydroxide solution and, by monitoring the conductivity of this solution, the oxygen content of the compound was calculated. Salzer  used a pure carbon filling (without platinum) for reduction at 1220°C. The carbon monoxide was then oxidized to carbon dioxide with iodine pentoxide and the gas stream was conducted into a 0.018 N solution of potassium hydroxide and the change in conductivity was measured. A similar method is that described earlier by Malissa and Schmidts , but they used 0.005 N sodium hydroxide solution as the absorbent. Ehrenberger et al.  described a semi-automatic oxygen analyzer. The oxygen content of the products of pyrolysis were converted into carbon monoxide on a carbon layer at 1140°C, then oxidized on iodine pentoxide to carbon dioxide, which was conducted into an absorbent solution containing barium hydroxide, hydrogen peroxide and ethanol. The p H of the solution was monitored. Boos  was the first to used a thermal conductivity cell for the determination of oxygen in organic compounds. The combustion products were conducted with helium carrier gas to a carbon layer kept at 1200°C. The gas mixture was purified by means of appropriate absorbent and a 13X molecular sieve, and the gas was then led into a thermal conductivity cell. The result was obtained from the chromatogram. Gotz  described a similar method. For the determination of elements other than carbon, hydrogen, nitrogen and oxygen, automatic methods are rarely used, as they are not normally determined in large series. For the determination of sulphur conductimetric titrations or conductivity measurements are sometimes used in an automatic version. Kainz and Muller  developed an automatic method for the determination of halogens with which one determination can be completed in 4 min. The combustion is carried out in the expanding chamber of an empty combustion tube at 1000°C in a stream of oxygen. The combustion is completed in the extended narrow part of the combustion tube which 506
contains platinum at 800°C The halogens in the carrier gas form halide ions in the absorbent solution, which contains hydrogen sulphite. The titration of halides is carried out by the null-point potentiometric method using 0.01 N silver nitrate solution, a silver indicator electrode and a suitable reference electrode. The automatic titration is carried out with a 10-cm burette with a sample weight of 3-5 mg. 3
1. Automatic analyzer units We call automatic analyzers, as indicated in the introduction, single or combined apparatus (consisting of modular parts) which carry out the analysis of solids, liquids or gases completely automatically and produce the results directly or in the form of an easily evaluable output (e.g., a gas chromatogram). In this section we discuss the following two important types of automatic analyzer units: 1. Instruments suitable for the determination of the main elements of organic samples (mainly carbon, hydrogen, nitrogen and oxygen), by analyzing the gaseous products obtained by combustion. 2. Instruments suitable for the analysis of solutions prepared from dissolved or mineralized organic compounds. The (sometimes automatically) weighed sample is digested, converted chemically into suitable forms, and the measurement is made by spectrophotometry (or by another suitable method, e.g., electrometrically). Laboratories in the food industry (protein nitrogen), agricultural (nitrogen in animal feedstuffs, fertilizers), and medical laboratories (determination of different body fluids such as urine or blood) are the main users of these instruments. We d o not discuss several kinds of other devices, in which the treatment of the sample is not made by combustion or wet digestion, their role being separation only, and the separated components being determined in their original form (e.g., amino acid analysers). Similarly, we omit instruments suitable for the automatic determination of impurities in gases (e.g., airpollution control apparatus or industrial gas analyzers), as well as devices capable of determining dissolved, mostly ionic components of water or other liquids (e.g., automatic water hardness meters or chloride meters). These instruments may run intermittently, handling the samples separately, and stopping once the analysis is completed. On the other hand, there are continuous instruments which bring the sample into the apparatus
automatically after completing one determination. These work continuously, without outside intervention or regulation, and produce the results in a relatively short time. Automatic analyzer units belonging to the first group are mainly of the intermittent running type, these were described in the first part of this chapter. An important accessory of these is the electric microbalance (Cahn, Mettler, etc.), which weighs the sample (in the milligram range) automatically and displays the weight without the necessity for manual operations. These are nowadays standard accessories of the analyzer instruments. There are instruments  which take the sample, weigh it and transfer it into the digestion vessel. These are usually macro or semimicro devices. Merz  described a relatively simple instrument for the automatic determination of nitrogen. The sample, wrapped in aluminium or tin foil, is dropped into a vertical silica tube where it is combusted in a stream of pure oxygen. The oxygen flow is then interrupted and changed to carbon dioxide automatically. The gas is led through a copper oxide layer and then through a copper layer into an automatic azotometer. The process is controlled with an automatic programme device. The duration of one determination is 2.5-3 min. One type of automatic instrument, suitable for the determination of carbon, hydrogen, nitrogen and sometimes oxygen, separates the components of the gas mixture obtained by combustion on a gas chromatographic column. C a r b o n dioxide, water vapour (or another gas obtained from it by chemical reactions, e.g., acetylene) and nitrogen are eluted after each other with a suitable carrier gas and passed to the detector. Another type of instrument has more detector units, and the gas mixture obtained in the combustion apparatus, after mixing, is divided into proportional parts and conducted into separate detector units, the interfering components having been removed previously by suitable absorbers. The chromatogram drawn by the instruments can be evaluated by measuring the peak heights or peak areas. Often, the results are printed out in digital form. Mostly thermal conductivity detectors are used because of their simple and trouble-free operation, but for these a suitable carrier gas is necessary in which even small amounts of the investigated component cause measurable effects. Hydrogen would be best for this purpose, but because of the dangers associated with its manipulation, helium is preferred. It would be best to use specific detectors for each measured component. In those instruments which are suitable for the determination of oxygen it is necessary to change the combustion tube to a reaction tube which contains a carbon layer. This tube is operated at a higher temperature. There are also types of instruments in which the two parts used
for the carbon, hydrogen and nitrogen and for the oxygen determination are built in together. Merz  described an automatic method for the determination of oxygen in the presence of metals and in organometallic compounds. He found that for the separation of oxygen from several metal oxides a mixture of a m m o n i u m chloride, silver chloride and hexamethylenetetramine is suitable. His instrument produces carbon dioxide from the oxygen in the sample and, after absorption, titrates the mixture automatically using photometric end-point detection. Pella and Colombo  described an automatic device suitable for the determination of 100-300 |xg of oxygen by a pyrolysis gas chromatographic method. The investigated sample is pyrolysed in helium carrier gas, then the gas is led through a carbon layer at 1000°C, where carbon monoxide is formed. This is then separated from the other gases by gas chromatography. The determination is carried out with a thermal conductivity detector. If the sample contains chlorine, the carbon filling has to contain 15% of nickel and 30% of platinum. Figure 50 shows schematically an instrument suitable for the determination of carbon, hydrogen, nitrogen and oxygen . One part of the instrument (on the right) carries out the determination of carbon, hydrogen and nitrogen, while on the left the determination of oxygen takes place. Absorber tube 1 on the right removes the possible impurities in the helium-oxygen mixture. The sample is placed in the combustion tube of the small furnace 2, where the pyrolysis and combustion of the sample take place. Block I contains two reaction tubes. The first of these contains copper oxide at 960°C, which completes the oxidation of the pyrolysis products. The second tube, connected to the first, contains a relatively long copper filling which removes the excess of oxygen from the carrier gas. Behind this there is a silver layer, kept at 500°C, which binds the halogen and sulphur content. The copper filling reduces all nitrogen oxides. Column III is packed with C h r o m o s o r b 102 and separates water vapour from carbon dioxide. Nitrogen travels with the carrier gas to the thermal conductivity detector, followed by carbon dioxide and finally water vapour. The gas chromatographic peaks are recorded, as can be seen in Fig. 51 (1). An integrator and digital converter transform the detector signal into numerical data. The left-hand side of the instrument is designed for the purpose of oxygen determination. Helium carrier gas is first purified in absorber 1. Pyrolysis is carried out in furnace 2 and products of the pyrolysis together with the carrier gas reach the reactor tube, which contains carbon and is kept at 1120°C. Oxygen is converted into carbon monoxide. The gas mixture having gone through absorber 3, which contains a mixture of lithium hydroxide, 509
Thermal conductivity detector
Registration apparatus Fig. 50. Schematic diagram of an apparatus for the determination of carbon, hydrogen and nitrogen working with a gas chromatographic column / — G a s purifier; 2—pyrolysis
Fig. 51. Chromatograms recorded with the apparatus shown in Fig. 50. ; ~ c , H, n : 2—o
magnesium perchlorate and phosphorus pentoxide, passes into column II, packed with molecular sieve 5A, kept at 120°C. The gas now contains carbon monoxide, nitrogen and helium only, and gives rise to the type of chromatogram shown in Fig. 51 (2). O n e can distinguish the signals of nitrogen, hydrogen and methane as well, but these do not interfere in the determination. A digital converter is used. The instrument has to be calibrated with standards. An automatic sampler, suitable for feeding 24 samples into the device, is also available. According to the manufacturer's manual, one can carry out 23 carbon, hydrogen and nitrogen determinations 510
in 4 h and 23 oxygen determinations in 2 h. Using an automatic electric balance, one technician can carry out 40 carbon, hydrogen and nitrogen and the same number of oxygen determinations (or 69 carbon, hydrogen and nitrogen determinations alone) in one working day. The size of the sample is 0.5-1.5 mg, and the standard deviations calculated from 90 determinations are 0.3% for carbon, 0 . 1 % for hydrogen, 0.2% for nitrogen and 0.3% for oxygen . According to another description  one carbon, hydrogen and nitrogen determination can be carried out in 10 min and with a standard deviation of 0.3%, while the time required for one oxygen determination is 5 min. Figure 52 shows the layout of an instrument of another design. Combustion is carried out in tube 1. The sample is weighed into a platinum or ceramic boat and is pyrolysed in the combustion tube heated to 900°C. Helium is used as carrier gas, containing some oxygen. Both gases have to be purified beforehand. We can carry out the pyrolysis of the sample with programmed heating as well. The products of the pyrolysis, together with the carrier gas, pass through the fillings, made of platinum, copper oxide and silver. O n these the products of pyrolysis are oxidized and the silver filling removes the product formed from the halogen and sulphur content of the sample. Tube 2 contains first silver and then copper metal, kept at 650 C, which removes the excess of oxygen from the mixture. Nitrogen oxides are also reduced on the filling. F r o m this point, the gas mixture contains only water vapour, carbon dioxide and nitrogen. It is then passed into mixer 3, of known volume, from which a sampler piston takes an appropriate volume, according to the concentrations in question. To determine the a m o u n t of water (that is, hydrogen in the original sample), the gas mixture first passes through the first thermal conductivity detector and then to an absorber tube, which contains magnesium perchlorate to remove the water content of the gas, then reaches the second detector, which is differentially connected to the first. Thus, the signal of the second detector, after amplification and integration, is digitized pointing out the hydrogen content. The helium carrier gas, which now contains only carbon dioxide and nitrogen, reaches the second detector (6). Between the two gas detectors is an absorber tube containing soda-asbestos. Thus, these detectors measure the a m o u n t of carbon dioxide in the gas, and so the carbon content of the sample can be calculated. The helium carrier gas, which now contains only nitrogen, finally reaches the third detector. A single detector provides a signal proportional to the nitrogen content. The apparatus is also suitable for the determination of oxygen if the furnace and combustion tube (/) are changed to a high-temperature furnace, kept at about 1120°C, and a ceramic or silica combustion tube which contains granules of carbon. The carbon monoxide, C
» solenoid valves © detectors
Fig. 52. Apparatus for the determination of carbon, hydrogen and nitrogen working with a thermal conductivity detector (a) C o m b u s t i o n s e c t i o n ; (b) m e a s u r i n g s e c t i o n ; / — t u b e furnace for c o m b u s t i o n of t h e s a m p l e in oxygen mixed in the helium carrier g a s ; 2 — t u b e furnace with p a c k i n g t o r e m o v e interfering o x i d a t i o n p r o d u c t s of excess o x y g e n a n d h e t e r o e l e m e n t s in the gas m i x t u r e ; 3 — h o m o g e n i z a t i o n c h a m b e r for t h e the gas m i x t u r e ; 4—apparatus m i x t u r e ; 5 — d e t e c t o r s for w a t e r ( h y d r o g e n ) ; 6—detectors
for t a k i n g k n o w n a n d identical a m o u n t from t h e gas
for c a r b o n d i o x i d e ; 7 — d e t e c t o r s for n i t r o g e n ; 8—absorption
p a c k e d with m a g n e s i u m p e r c h l o r a t e ; 9 — a b s o r p t i o n t u b e p a c k e d with s o d i u m a s b e s t o s for r e m o v i n g c a r b o n d i o x i d e
formed from the oxygen content, is converted into carbon dioxide on a filling of copper oxide. Once the sample has been placed in the combustion tube, the instrument works automatically, including the regulation of the temperatures of the furnaces as well as opening and closing of gas taps, etc. The instrument works with relatively small samples (about 1 mg), for the weighing of which a Cahn electric microbalance is most suitable, although other kinds of precise microbalances are also applicable. It is best to house the instrument in a separate room, protected from the laboratory atmosphere, as corrosion might affect the delicate mechanical parts. The time required for one determination of carbon, hydrogen and nitrogen is 15-20 min, but one cannot expect more than 20-25 determinations within a working day of 8 h, as at the beginning of the day the instrument has to be warmed u p and calibrated, which takes 1-2 h, especially if samples of different composition have to be analyzed. The instrument can be better utilized if it is run in 2 or 3 shifts per day. Maintenance is also necessary from time to time, e.g., to change the fillings of the various reaction tubes. There are several other instruments on the market, including the Haereus Ultramat C, H, N analyzer, in which the water is separated from the carbon dioxide and nitrogen by freezing. The separated components are detected with a thermal conductivity cell. The instrument, connected to a Cahn 512
automatic microbalance, is completely automatic. A detailed description is given in the book by Ehrenberger and Gorbach  (pp. 76-81). Various considerations have to be taken into account before deciding on the necessity for the purchase of an automatic analyzer unit. Its use is mostly recommended if a number of samples of not too different composition are to be analyzed on a routine basis. They are less handy for liquid samples, as the difficulties connected with the lengthy sample preparation (weighing into a capillary) might counterbalance the advantages. The instrument does not replace a well equipped microanalytical laboratory, because beside the need to determine carbon, hydrogen, nitrogen and oxygen routinely, the determination of other elements (e.g., halogens, sulphur) and functional groups will be necessary from time to time. An instrument can be run by one skilled technician, so there might be savings in manpower. It is very important that a trained technician is available, who knows the instrument and is able to carry out repair and maintenance work immediately. Of more recent papers, we review some that deal specifically with automatic elemental analysis. Kainz et al.  combusted the organic compound in a flow of oxygen at 900°C and the water vapour was removed by an absorber tube containing calcium chloride. The carbon dioxide content was led into a mixing chamber which contained 0.01 N sodium hydroxide solution and 2% ethanolamine, and from the change in the electrical conductivity of the liquid the carbon content was determined. The calcium chloride absorber was then heated to 350°C and the water removed was conducted into another mixing-chamber, which contained 2% sulphuric acid in glacial acetic acid. The electrical conductivity was again measured, which changes under the influence of a small amount of water. Merz  described a completely automatic instrument for carbon and hydrogen determination in which the combustion of organic compounds is carried out in a vertical silica tube in a stream of oxygen with the use of manganese dioxide. The water is frozen out of the gas flow and the carbon dioxide is absorbed in a titration cell, which contains dimethylformamide and an amine in solution, and the absorbed carbon dioxide is then titrated with tributylmethylammonium hydroxide using automatic colorimetric end-point detection. The evaporated water then is led through a carbon layer kept at 1120°C, where the water forms carbon monoxide, followed by oxidation to carbon dioxide. This is determined by the same method as described above. The instrument prints out the results of the analyses. If two titration cells are used, the time required for the determination of carbon and hydrogen together is 8 min.
Wachsberger et al.  described an automatic analyzer in which the combustion of organic c o m p o u n d s is carried out in a stream of oxygen and the combustion products mixed with helium, are separated by gas chromatography using a thermal conductivity cell. The detector signal is integrated and the results are finally printed out as percentages. The precision of the determination is the same as that of chemical methods. Stoffer  published results obtained from the automatic analysis of 23 organic compounds using a Carlo Erba Model 1102 automatic elemental analyzer, a Sartorius 4125 microbalance and a Combitron = S = 1 0 / 1 0 computer. This "off—on-line" automatic method gave reliable results. Salzer  wrote a critical review of physico-chemical methods that are applied in automatic carbon, hydrogen, nitrogen and oxygen analyzers.
2. Automatic wet digestion systems In the last 15-20 years the increased workload of agricultural, food, industrial and hospital laboratories has necessitated the development of analytical methods that are suitable for the serial determination of a single component in samples of similar composition on the semimicro- or microscale. Even such a well developed classical analytical method as the determination of nitrogen in proteins by the Kjeldahl method does not allow one person to carry out more than 3-4 determinations per hour, even if 20-30 digestions are carried out simultaneously. The time required for the distillation of the ammonia liberated after digestion is 10-15 min. Attempts to automate the Kjeldahl method have been made, in which the digestion of the organic compound as well as distillation and titration are carried out automatically. One instrument capable of this is the " K j e l - F o s s - M a k r o - A u t o m a t e " [53, 54], which has six digestion flasks. The digestion with sulphuric acid is aided with tablets which contain a catalyst and a compound which elevates the boiling point. The digestion is accelerated by the automatic addition of hydrogen peroxide, the a m o u n t of which is automatically increased if the investigated sample contains more lipids. The optimal temperature of the digestion is 410°C. After this, the contents of the flask are cooled quickly and, after automatic dilution, an automatic burette adds the necessary alkali to the solution, and the distillation is automatically started. Another automatic burette titrates the distilled liquid continuously with standard sulphuric acid and the change in colour of the indicator at the end-point is detected with a photoelectric device. The instrument prints out the results. Subsequently the 514
flask is emptied automatically and a new determination can be started. The sample weight is 0.5-1.0 g, depending on the protein content of the sample. According to the manufacturer's instructions, the instrument is capable of analyzing 20 samples per hour; the time required for a single analysis is 12 min. The advantage of the method over others described later is that the ammonia, distilled automatically, is titrated by alkalimetric method. These classical methods are very reliable, whereas spectrophotometric methods for the determination of ammonia, because of their too high sensitivity, are less reliable. The application of the apparatus is also advantageous because of the relatively large samples they use, which reduces the danger of sampling errors. In the last decade continuous automatic analyzers had been developed that are suitable for the determination not only of nitrogen, but also of other elements such as phosphorus, potassium, calcium, aluminium and halogens. The development of these instruments was made possible first of all by the availability of peristaltic pumps. The most important part of a peristaltic p u m p is a thin, flexible, plastic tube which is compressed with a forwardmoving cylinder of a rotating disk along the tube and is released after travelling a given distance. Because of this the liquid moves at a uniform rate inside the tube. At pre-determined intervals air (or another gas, sometimes a liquid) is fed into the tube, so that the liquid portions are separated and these move forward practically without diffusion or mixing. The individual liquid portions enter one after the other, separated with an air column of predetermined length, the mixer and reaction chambers or coils to which automatical pumps feed the necessary diluent, buffer and reagent solutions. Also, the temperature of the reaction system can be regulated. The solutions may even pass into a pre-heated tube system to allow the necessary time to pass to make a reaction complete, e.g., development of a colour. Subsequently, if the mixture is coloured, the solution flows into the flowthrough cell of a colorimeter or spectrophotometer. O n e of the two light beams (of equal intensity) passes through the sample cell while the other passes through the reference cell, containing the solvent or a blank. The two light beams fall ultimately on two photoelectric cells which are switched in a bridge circuit, their net current difference being amplified, digitized and finally printed out. If the instrument is calibrated with standard samples one can obtain the concentration directly. An important part of an automatic wet digestion system is the rotating digestion tube (helix) in which the Kjeldahl-type digestions can be carried out continuously (Fig. 53). The measured aliquots of the samples, after mixing with the digestion reagent, flow one after the other into the rotating helical tube, which is heated from the outside. Each sample is separated from the next 35
Fig. 53. Rotating digestion tube (helix) in the automatic nitrogen analyzer (a) Section with lower t e m p e r a t u r e ; (b) section with h i g h e r t e m p e r a t u r e ; / — m i x t u r e of t h e s a m p l e a n d t h e a c i d ; 2—liquid level; 3—digestion p r o d u c t solution
with larger air bubbles or portions of an indifferent liquid. Each liquid sample remains inside the tube, moving slowly forwards, until digestion is complete. The temperature of the different parts of the digestion tube can be regulated separately. Usually the digestion is started at a higher temperature, but the sample remains longer in the lower temperature part of the tube. The method has the merit of being continuous and rapid, the system is protected from impurities, and the results are obtained without h u m a n error. The instrument works with small volumes and requires small amounts of reagents. As the instrument provides relative results (relative to standard samples), usually the error of the determination depends only on the accuracy of the last measurement taken in the system. In spectrophotometric or colorimetric measurements, the error is about 5%, but it is lower if other methods, e.g., electrometric, are used. For the calibration and continuous monitoring of the process a pure reagent (e.g., for the determination of nitrogen, a m m o n i u m sulphate) is usually suitable, but most often it is necessary that the standard samples have almost the same composition as the investigated samples. The Technicon AutoAnalyzer is the best known and most generally used of these instruments. The instrument is designed on the modular principle, and an optimum combination of parts can be assembled for a given analytical task. A schematic diagram of a combination assembled for the determination of nitrogen is shown on Fig. 54. Part (a) consists of the sampler and digester unit while part (b) is the analyzer itself. In the holes of the periodically rotating plate of the sampler (3) are placed dishes which contain the liquid samples, from where an automatic sampler draws u p a pre-determined volume of the sample. Usually it draws 20-30 samples per hour. Solid samples can also be analyzed if the sampler is connected to a special automatic balance (Fig. 55) which weighs the sample and moves it from the balance plate to the dissolution apparatus, and feeds the corresponding electrical signal from the weighed sample into a microprocessor unit. A peristaltic p u m p brings the liquid samples (separated from each other with air bubbles) into the digestion tube. One or two pumps add the digestion reagents to the sample. After digestion, the sample is diluted with water and the solution is moved with 516
Sampler: 20 sample .per hour
to the sampler ^ 2.00 cm /min water - — ®0.32 cm /min air 3
Biamperometric d e measuring cell bubbler to the pump
0.40 cm /min sample sol 3
~f~^ 0.23 cm /min KI-solution 3
^ 0.80 cm /min HCl-solution 3
to the waste
|Recorder) Fig. 54. Diagram of the apparatus design for the determination of nitrogen ("AutoAnalyzer") (a) S a m p l e r a n d d i g e s t o r ; (b) a n a l y z e r ; / a n d 2 — p u m p s y s t e m s ; 3 — r o t a t i n g plate
Fig. 55. Schematic diagram of an automatic analytical balance for weighing solid organic samples / — B a l a n c e ; 2—balance
p l a t e ; 3—feed-in
w e i g h e d ; 5—excess s a m p l e ; 6—cover plate
another p u m p of p u m p system 2. The solution which left the digestion tube reaches the mixer, from where a suitable aliquot is moved into the analyzer system with one of the pumps of p u m p system 1. The excess of liquid remaining in the mixer is removed with another p u m p before the next digested sample enters. A separate p u m p injects sodium hydroxide solution 35*
into the analyzer to neutralize the solution and render it alkaline. This takes place in the first mixer. When the liquid leaves the first mixer, sodium phenolate solution is added, is mixed with the liquid in the second mixer. Finally, a p u m p adds sodium hypochlorite solution to the solution, which now contains all of the necessary reagents for colour formation. It is then moved into a helical tube kept at 90°C and, when it leaves this tube, the colour has the highest intensity. The indophenol blue type compound obtained, formed from ammonium ion, sodium hypochlorite and sodium phenolate (and the necessary buffer solution and catalyst), is characterized with the structure shown on p. 334. The mechanism of the colour formation reaction is not clear. There are more parallel oxidation and chlorination processes and the relative rates of these depend on the conditions of the reaction. So we can use this colour formation reaction for determination of ammonia only in a mechanized system, where the conditions of the reaction are the same every time. The coloured solution passes through a cuvet of the spectrophotometer and the signal of the instrument draws a diagram in the registrator apparatus. By a similar way — the other part of the instrument is combined — we can carry out another analytical determination. We can connect a dialyser to the guide necessary for the separation of low molecular weight compounds from the proteins or other colloids. Instead of a spectrophotometer we can use a flame photometer for the determination of alkali- and alkali-earth metals Atomic absorption instrument, coulometric measuring cell, pH-meter or ionselective electrode measures directly the concentration of ions. We can see the combination of determination of nitrite ions in Fig. 56. The nitrite ions Recorder
Fig. 56. Schematic diagram of an automated nitrite determination apparatus
liberate iodine from a iodide content solution, which we can determine very sensitively with a biamperometric instrument. We can find several papers in the scientific journals of the last 15 years, on the development and application of automated analytical methods. Ferrari  proposed an automatic method for the determination of protein nitrogen in a Kjeldahl flask. Cedergreen and Johansen  also digested the sample in a Kjeldahl flask, which took 25 min, then oxidized the ammonia in alkaline solution with bromine generated coulometrically at the anode: 2 B r " - B r + 2e 2
Br + 4 0 H 2
+2H 0 2
+2NH ->N + 3 B r + 3 H 0 3
Using this method, as there was no distillation, the time of the analysis was shortened considerably and they could carry out 15 determinations per hour. The standard deviation was 0.2% and they could determine as little as 0.5 jig of nitrogen in organic compounds. Marten and Catanzaro  automated the Kjeldahl nitrogen determination and investigated the effects of digestion reagents, different catalysts and the temperature of digestion. They used a special digestion tube in which they could even determine the nitrogen content of cyclic nitrogen compounds (e.g., nicotinamide) with 9 9 . 1 % efficiency. K r a m e et al.  described a modified apparatus for the determination of a m m o n i u m ions in which, with the proper choice of the digestion and colour-forming reagents, the phenol hypochlorite reaction became much more sensitive.. Using their digestion method it was not necessary to use a standard of similar composition to the sample as they obtained good results with pure ammonium sulphate. Their digestion reagent consisted of 200 c m of 1 + 1 (v/v) perchloric acid, 3 g of selenium dioxide and 1000 c m of concentrated sulphuric acid. They did not observe any loss of nitrogen even if the carbohydrate content of the sample was more than 10%. 3
Automatic analyzers, especially of the Technicon type, have been used for the determination of elements other than nitrogen. For example, Hofstader  determined phosphorus in addition to nitrogen. After digestion the nitrogen was determined spectrophotometrically using the phenol hypochlorite reaction, and 30 determinations of nitrogen could be carried out per hour. T o determine phosphorus the organic compound was automatically digested with a mixture of sulphuric, nitric and perchloric acids and, after suitable treatment, phosphorus was determined colorimetrically using the molybdophosphovanadate method. 519
Ruzicka and co-workers [60, 61] published a method based on flow injection analysis. About 0.5-cm volumes of sample solution were injected with a syringe into a quickly flowing stream of carrier liquid (reagent) in rapid succession, so that they could complete 400-700 determinations per hour. The method was used mainly for the determination of phosphorus. Varley  determined the nitrogen, phosphorus and potassium content of plant samples with the Technicon instrument. After acidic digestion, ammonia was measured spectrophotometrically using the phenol hypochlorite reaction. According to the author, the other method suggested for this purpose, the so-called "bias" method, is unnecessarily oversensitive. The phosphorus content was also determined by spectrophotometry using the yellow colour of the molybdophosphovanadate complex. The determination of potassium was carried out on an aliquot of the digested solution by a flamephotometric method using lithium nitrate as the internal standard. Forty nitrogen, 60 phosphorus and 60 potassium determinations could be carried out per hour. Docherty  described an automatic sampling and analysis apparatus for the serial analysis of fertilizers. The instrument was connected to a Mettler Model D W A C30 automatic balance, from which the weighed sample moved to the dissolution, dilution and analysis units. The ammonium ion content of the fertilizer was determined by means of the phenol hypochlorite reaction, the phosphorus content by utilizing the molybdophosphovanadate complex and the nitrate content by the 2,4-xylenol reaction. The determination of potassium was carried out by flame photometry. In another paper Docherty  described a method for the determination of potassium, phosphorus and ammonium-nitrogen in fertilizers. The automatic method of Roach  for the determination of calcium and phosphorus is also available. The spectrophotometric determination of calcium was carried out with cresolphthalein (2,6-xylenolphthalein-a,a'-bisiminodiacetic acid) at p H 10.7 at 580 nm using an 8-mm flow-through cell. For the determination of phosphate the molybdophosphovanadate method was used. The water analysis apparatus of Britt  is suitable for the determination of chloride, nitrate, nitrite and iron (III) ions in water at the p p b level. Weinstein et al.  developed a semi-automatic method for the determination of fluorine in air and plant materials. Bieder and Brunei  described the simultaneous determination of iron(II), iron(III) and nitrogen in pharmaceutical products using the Technicon AutoAnalyzer. Dabin  described the determination of several elements, e.g., nitrogen, carbon, phosphorus, aluminium, iron and titanium, with the Technicon AutoAnalyzer. Wrightman and McCadden  developed an automatic method for aluminium, based on the fact that aluminium liberates equivalent amounts of iron from the iron(III) E D T A complex. Iron can then be titrated in the presence of 2,4,6-tripyridyl-s-triazine 3
indicator. Holl et al.  described an automatic spectrophotometric method for the analysis of pharmaceutical products. They prepared solutions from single tablets with water, 0.1 N sulphuric acid and 0.1 N sodium hydroxide solution and their UV absorbances were measured at 285, 280 and 295 nm, respectively. As the sample is not investigated chemically, the method is rapid and suitable for identification of the product. Ashbolt et al.  developed an automatic process in which they measured the concentration and the acidity of solutions of drugs with two AutoAnalyzers in parallel. Foster  described in a review the possibilities of pharmaceutical applications of automatic analytical techniques. Michaels et al.  used the Technicon AutoAnalyzer for the investigation of the rate of dissolution of pharmaceutical tablets. Sodergreen  determined surface-active agents in water with an automatic apparatus. F o r the spectrophotometric determination of alkylbenzene sulphomates the Methylene Blue method was used. 14 samples per hour could be analyzed. Several workers investigated the possibility of the automatic determination of carbohydrates. Of spectrophotometric methods, the oxidation of glucose with hexacyanoferrate(III) is suitable, the decrease in absorbance of the coloured solution being measured. Of electrometric methods, the measurement of the oxidation-reduction potential of the hexacyanoferrate(III)/(II) system is suitable for the determination of glucose. Porter and Sawyer  used oxidation with hexacyanoferrate(III) for the determination of carbohydrates, after hydrolysis in foods, the change in potential being monitored with a redox detector. Thirty samples per hour could be analysed. The method is suitable for the determination of 0.0025-0.5% of dextrose content. Llenado and Rechnitz  described an automatic enzymatic method for the determination of glucose in serum solutions, based on the following reactions: ^ ~ Glucose + H 0 + 0 2
H 0 + 2I+2H 2
glucose oxidase amine — -----------• Mo(VI) catalyst
. ., _ gluconic-acid + H 0 IW
, ^ _ I + 2H 0 2
The reaction was monitored with an ion-selective electrode. Seventy determinations per hour could be carried out. Marten  described an apparatus suitable for the automatic analysis of pharmaceutical products, fertilizers, foodstuffs and synthetic detergents. Ferrari et al.  used an automatic method for the determination of penicillin and streptomycin in fermentation liquids. The separation of antibiotics from the liquid was carried out with a double dialyser; for the determination of penicillin an iodimetric method was used, while the 521
determination of streptomycin was carried out spectrophotometrically with iron(III) chloride reagent. The error of the method is 3%. Only very few papers have dealt with the automatic determination of functional groups. A possible explanation for this is that serial determinations of functional groups are rarely needed so although there are several possibilities, the construction of automatic instruments is not necessary. If there is a need, one can carry out serial determinations with a gas chromatograph or mass spectrometer with manual sample injection. Technicon have patented a method for the determination of the alkoxy group . Less than 1 mg of sample is heated with concentrated hydroiodic acid and the alkyl iodides formed are transferred with helium carrier gas into a thermal conductivity detector for quantitative determination. Foreman and Stockwell  have discussed the problems of automatic analysis in depth. Ehrenberger and Gorbach  discussed different concepts of automation and different automatic analysis systems, and also gave useful practical data on their applications. In the author's opinion, future developments will be concentrated on the use of mass spectrometry, which will make possible the determination of functional groups also. Possibly, neutron activation analysis will be automated, mainly for the determination of trace amounts of impurities. High-resolution spectrophotometric methods will also have a role. The costs of these instruments and other economic considerations are behind the observable trend that in many countries central microanalytical laboratories have been established, with well trained personnel and with all the expensive equipment, where non-routine analytical problems can be solved, and the results obtained within 8-24 h. Institutions, factories, etc., can therefore concentrate more on routine-type work. It is obvious,however, that with increasing automation and sophistication, the demand for well qualified analytical chemists will continue to rise.
References to Chapter 10 1. Trutnovsky, H.: Mikrochimica Acta, 909 (1971). 2. Korshun, M. O. (Ed. J. Kuck): Methods in Microanalysis, Simultaneous Rapid Combustion. Gordon and Breach, New York, 1964, Vol. 1. 3. Robertson, G. I., Jett, L. M., Dorfman, L.: Anal. Chem., 30, 132 (1958). 4. Robertson, G. I., Jett, L. M., Dorfman, L.: Anal. Chem., 32, 1721 (1960). 5. Gel'man, N. E., Bresler, P. I., Ruzin, B. N., Grek, N. V., Sheveleva, N. S., Melnikov, A. A.: Dokl. Akad. Nauk SSSR, 161, 10 (1965). 6. Nail, W. R., Schooley, R.: Metallurgia, 64, 9 (1961). 7. Vecera, M., Lakomy, J., Lehar, L.: Mikrochimica Acta, 674 (1965). 8. Stuck, W.: Microchem. J., 10, 202 (1966).
9. 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.
Malissa, H.: Mikrochimica Acta, 127 (1960). Malissa, H., Schmidts, W.: Microchem. J., 8, 180 (1964). Pell, E., Macherdl, L., Malissa, H.: Microchem. J., 10, 2867 (1966). Stuck, W.: Mikrochimica Acta. 421 (1960). Greenfield, S.: Analyst, 85, 486 (1960). Greenfield, S., Smith, R. A. D.: Analyst, 87, 875 (1962). Greenfield, S., Smith, R. A. D.: Analyst, 88, 886 (1963). Gouverneur, P., Van Leuven, H. C. E., Belcher, R., Macdonald, A. M. G.: Anal. Chim. Acta, 30, 328 (1964). Gouverneur, P., Van Leuven. H. C. E., Belcher, R., Macdonald, A. M. G.: Anal. Chim. Acta. 33, 360 (1965). Keidel, F. A.: Anal. Chem., 31. 2043 (1959). Salzer, F.: Z. anal. Chem., 205, 66 (1964). Salzer, F.: Microchem. J., 10, 27 (1966). Haber, H. S., Gardiner. K. W.: Microchem. J., 6, 83 (1962). Haber, H. S., Bude, D. A., Buck, R. P., Gardiner, K. W.: Anal. Chem., 37, 116 (1965). Olson, E. C , Houtman, R. L., Struck, W. A.: Microchem. J., 5, 11 (1961). Francis, H. J. Jr.: Anal. Chem.. 36, 31 A-47 A (1964). Schoniger, W.. Pure and Applied Chem., 21, 497 (1970). Duswall, A. A., Brandt, W. W.: Anal. Chem., 32, 272 (1960). Sundberg, O. E., Maresh, C : Anal. Chem., 32, 274 (1960). Vogel, A. M., Quattrone, J. J.: Anal. Chem., 32, 1754 (1960). Walish, W:Ber. Deut. Chem. Ges., 94, 2314 (1961). Sommer, P. F., Gauter, W., Simon, W.: Helv. Chim. Acta, 45, 595 (1962). Miller, C. D., Winefordner, J. D.: Microchem. J., 8, 334 (1964). Pella, E., Colombo, B.: Anal. Chem., 44, 1563 (1972). Pella, E., Colombo, B.: Mikrochimica Acta, 698 (1973). Gorbach, S., Ehrenberger, F.: Z. anal. Chem., 181, 100 (1961). Gel'man, N. E., Wang Wen-Yun, Bryshkova, I. 1.: Zavods. Lab., 27, 24 (1961). Salzer, F.: Mikrochimica Acta, 835 (1962). Malissa, H., Schmidts, W.: Microchim. J., 8, 180 (1964). Ehrenberger, F.. Gorbach, S., Mann, U.: Mikrochimica Acta, 778 (1958). Boos, R. N.: Microchem. J.. 6, 389 (1964). Gotz, A.: Z. anal. Chem., 181, 92 (1961). Kainz, G., Muller, H. A.: Z. anal. Chem., 253, 180 (1971). Poy, F.: Chem. Rdsch., 12, 215 (1970). Merz, W.: Z. anal. Chem., 237, 272 (1968). Merz, W.: Anal. Chim. Acta, 50, 305 (1970). Pella, E., Colombo, B.: Anal. Chem., 44, 1563 (1972). Poy, F.: Carlo Erba, Milano (1972). Carlo Erba, Prospectus of C H N + O Elementar Analyser Model 1102. Kainz, G., Zidek, K., Chromy, G.: Mikrochimica Acta, 235 (1968). Merz, W.: Anal. Chim. Acta, 48, 381 (1969). Wachsberger, G., Dirscherl, A., Pulver, K.: Microchem. J., 16, 318 (1971). Stoffer, R.: Mikrochimica Acta, 242 (1972). Salzer, F.: Microchem. J., 16, 145 (1971). Banyai, E., Gimesi, O., Lendvay, Zs.: Periodica Polytechnica, 20, 115 (1976). Banyai, E., Gimesi, O., Lendvay, Zs.: Periodica Polytechnica, 20, 118 (1976). Ferrari, A.: Ann. N. Y. Akad. Sci., 87, 792 (1960). Cedergreen, A., Johansen, G.: Science Tools, The LKB Instrument Journal, 16, 2, 19 (1969).
57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.
Marten, J. F., Catanzaro, G.: Analyst, 91, 42 (1960). Krame, D . G., Griffen, R. H., Hartford, C. G., Corrado, J. A.: Anal. Chem., 45, 405 (1973). Hofstader, R. A.: Microchem. J., 10, 444 (1966). Ruzicka, J,'Hansen, E. H.: Anal. Chim. Acta, 78, 145 (1975). Ruzicka, J, Stewart, J. W. B.: Anal. Chim. Acta, 79, 79 (1975). Varley, J. A.: Analyst, 91, 119 (1966). Docherty, A. C : Technicon Symposium: Automation in Anal. Chem., N e w York, 1976. Docherty, A. C : Imperial Chemical Industries Inc. N e w York (1968). Roach, A. G.: Technicon 5. Int. Symp.: Automation in Anal. Chem., N e w York (1965). Britt, R. D.: Anal. Chem., 34, 1728 (1962). Weinstein, L. H., Mandl, R. H., McCune, D. C , Jay, S. J, Hitchcock, A. E.: Boyce Thompson Inst, for Plant Research Inc., 22 (4) 207 (1963). Bieder, A., Brunei, P.: Ann. of the New York Acad, of Sci., 130, 627 (1965). Dabin, B.: Symp. 1965-Technicon, France Wrightman, K. B., McCadden, R. F.: Ann. of the New York Acad., of Sci., 130, 827 (1965). Holl, W. W., Tukefjian, J. H., Michaels, Th. P., Sinotte, L. P.: Ann. of the New York Acad, of Sci., 130 (2) 525 (1965). Ashbolt, R. F., Bishop, D. W., Styles, E. E.: Ann. of the New York Acad, of Sci., 130, 835 (1965). Foster, M. C.: Automatic Analysis in the Pharmaceutical Industry. The Australian Journal of Pharmacy, (1976). Michaels, Th., Greely, V. J., Holl, W. W., Sinotte, L. P.: Ann. of the New York Acad, of Sci., 130, 568 (1965). Sodergreen, A.: Analyst, 91, 113 (1966). Porter, D . G., Sawyer, R.: Analyst, 97, 569 (1972). Llenado, R. A., Rechnitz, G. A.: Anal. Chem., 45, 2165 (1973). Marten, J. F.: Automated Preparation of Solids for the AutoAnalyser Technicon AutoAnalyser, Technicon Co. N e w York, 1967. Ferrari, A., Russo-Alesi, F. M., Kelly, J. M.: Anal. Chem., 31, 1710 (1959). British Patent 1038 703 (12.2.64) Technicon Instr. Corp; Ref., Anal. Abstr., 14, 195 (1967). Foreman, J. K., Stock well, P. B.: Automatic Chemical Analysis, Wiley, New York, 1975. Ehrenberger, F., Gorbach, S.: Methoden der organischen Elementar- und Spurenanalyse. Verlag Chemie, Berlin, 1973, pp. 40-93.