REPORT
s Ρ •D H l·-
(Q
s § ι •
S Efl
Classical Analysis A Look at the (Past, (Presentj and future
D
espite the fact that instrumental analysis hasrightfullyassumed an overwhelmingly major role in the analytical laboratory, there remains a limited, although important, need for clas sical analysis. Instrumental analysis is most useful for elemental determinations at minor and trace levels (about 1% all the way down to 1 atom), and in this range classical analysis performs either poorly or not at all. However, instrumental analy sis generally does not give high precision and accuracy at major levels (about 1% up to 100%), and in this range classical analy sis does perform well. Moreover, instru mental and classical analysis complement each other and can be used in tandem to the analyst's advantage. In addition, classi224 A
analysis will always be needed in order Although classical calto establish standards for instrumental calibration. analysis has taken The purpose of this article is to define and discuss the scope of classical analysis a back seat to and to explore die importance of renew the chemical community's interest in instrumental analysis ing it The histories of gravimetry and titrimeare traced, including biographical since the 1940s, its try sketches of some important personalities the development of classical analysis in revival is vital to the inEurope and the United States. The signifi cance of the events surrounding the Karls interests of industry ruhe Congress of 1860 and the impor Charles M. B e c k II National Institute of Standards and Technology
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994
tance of physical chemistry and organic reagents to the development of classical analysis are discussed. The current crisis in the United States resulting from a This article not subject to U.S. copyright. Published 1994 American Chemical Society.
shortage of qualified classical analysts is examined. The future use of classical and instrumental analysis in tandem is examined. Because of the serious economic consequences to American industry and government that would result from the disappearance of classical analysis, a proposal is made for the renewal of education in this field. What is classical analysis?
If the final step of an analysis is a gravimetric ortitrimetricdetermination, it is considered a classical analysis. The following test may be useful in distinguishing classical from instrumental analyses. The calculations for a classical analysis require no more than experimentally measured weights (or weights and volumes), definite chemical reactions, and atomic weights. Such a test demonstrates, for example, that a potentiometrictitrationis a classical analysis, because the potentiometer serves only to indicate the endpoint of thetitration.It also demonstrates that a spectrophotometric determination is an instrumental analysis because the calculations require an experimentally measured transmittance of light through a solution. There are two good reasons for clearly distinguishing between a classical and an instrumental analysis. First, in the hands of an experienced analyst, the relative precision of a classical analysis is about 0.1%0.2% or better, whereas the relative precision of an instrumental analysis is about l%-2% (although certain exceptions exist). Second, whereas almost all instrumental analyses are comparative and require known standards for calibration, classical analyses do not require external standards for calibration if the stoichiometries implied by the written chemical reactions reflect what is actually happening in the analytical process. For gravimetry this implies that actual precipitates correspond to written formulas, and fortitrimetrythat actual reactions correspond to written reactions and that indicators detect true equivalence points. We have advanced far enough beyond the 1940s to see that that decade was a watershed period marking the end of the
development and general use of classical analysis and the beginning of instrumental analysis. Although the use of instrumental analysis had been growing slowly up until thattime,most samples were analyzed classically. Since the 1940s instrumental analysis has assumed more importance because of the ever-increasing demands for sensitivity, speed, and economy. (For a sketch of the history of instrumental analysis, see Reference 1.) After World War II, the existing, underutilized instrumentation, as well as subsequently developed instrumentation, was made practical with the introduction of the photomultiplier tube, the transistor, and the microprocessor. We now have sensitivities and speeds that would have been unimaginable even a few years ago, but, as has always been the case, new demands keep pushing the definitions of "fast" and "ultratrace" to new levels. Because many, if not most, analytical chemists in the work force today were educated after the beginning of the instrumental era in the 1950s, it is useful to review the history of classical analysis. An awareness of the history of a discipline helps to provide a perspective for évaluât-
Chemistry is at least as oCd as recorded history, but whatzue recognize as eiqienmentat chemistry did not emerge untiC the end of the sixteenth century.
ing current conditions and needs and for anticipating future possibilities. History of gravimetry to the 1850s
Gravimetry is the determination of an element through the measurement of the weight of an insoluble product of a definite chemical reaction involving that element. Tracing the history of gravimetry amounts to tracing the early history of chemistry. Chemistry is at least as old as recorded history, but what we recognize as experimental chemistry did not emerge until the end of the sixteenth century. Later, with the publication of his book The Sceptical Chymist in 1661, Robert Boyle began the process of putting chemistry on a sound scientific footing. Boyle firmly dismissed Aristotle's four elements of fire, air, water, and earth and Paracelsus' three principles of mercury, sulfur, and salt. Instead, he advocated a comprehensive experimental approach before attempting any theoretical statements. As the influence of Paracelsus waned, interest in chemistry shifted from medicine to mineralogy and metallurgy. However, chemists were influenced for another century by the "phlogiston theory," first advocated in 1681 by Johann Bêcher and popularized by Georg Ernst Stahl. According to this theory, when a metal burns or rusts it gives off a substance called phlogiston. When the phlogistonists observed that iron gained weight when it rusted, they simply postulated that phlogiston had negative weight. As damaging as this theory was to the progress of chemistry, it did provide a general concept, and testing that concept led to a study of chemical analysis and simple chemical reactions. Among those making such studies were the gas chemists Joseph Black, Henry Cavendish, Carl Scheele, Daniel Rutherford, and Joseph Priestley. Although Priestley was one of the last staunch defenders of the phlogiston theory, he initiated its downfall in 1774 when he isolated a gas by heating mercuric oxide in a closed system. He gave the name "dephlogisticated air" to the oxygen he collected. A few years earlier, Rutherford
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994 225 A
REPORT
and Cavendish had independently discovered nitrogen, which Rutherford called "phlogisticated air." Aware of Rutherford's and Priestley's work, Antoine Laurent Lavoisier dealt the death blow to the phlogiston theory by performing quantitative experiments with mercury and air in a closed system. He correctly explained combustion and demonstrated that air was a mixture of nitrogen and oxygen. His genius, his instinctive recognition of the law of conservation of mass, and his use of the balance made him the forefather of the quantitative era in chemistry. At about the same time, the Swedish chemist Torbern Bergman outlined a systematic scheme of qualitative and quantitative analysis and described all of the available known reagents. His gravimetric method for silica is recognizable as a direct forerunner of the one we use today. Strangely enough, Bergman remained a firm believer in the phlogiston theory all his life. Although titrimetry already was being used in France at the turn of the nineteenth century, almost all analyses were done by gravimetry. For example, Sigismund Andreas Marggraf had worked out the determination of silver as silver chloride as a substitute method for fire assay which, under certain conditions, tended to give low results for silver. (Fire assay is the ancient art of isolating precious metals from an ore. The method is based on a high-temperature liquid-liquid extraction in which the ore is fused with a mixture of lead oxide, flux, and a reducing agent. Molten lead is formed throughout as the mixture is heated and the precious metals pass into the lead. Upon cooling, a lead "button" forms at the bottom of the mass from which the precious metals can be extracted.) Bergman was already using gravimetric factors to calculate the amounts of iron, lead, copper, and silver in various precipitates, and others were becoming interested in the analysis of minerals and industrial materials. At the turn of the nineteenth century chemistry entered a period of great confusion. Gravimetry had been developing through the eighteenth century in an empirical manner, because the laws of chemical composition were not understood. Many chemists believed that substances combined in definite proportions. For ex226 A
ample, they knew that a certain weight of silver always gave the same weight of silver chloride. Jeremias Benjamin Richter, a mining engineer in Silesia, also believed that there was an equivalence inherent in chemical reactions. He tried to work out the mathematical relationships and coined the word "stoichiometry" for the proportions existing among various substances. Unfortunately, he was led down many a blind path because he tried to force his data to fit his preconceptions. His work was a step in the right direction, however, because he had an intuitive sense of the law of definite composition. Such was not the case with the famous French chemist Claude Louis Berthollet, who wrongly postulated that the composition of a compound of two elements might vary between maximum and minimum in all proportions. Joseph Louis Proust opposed Berthollet's view and presented experimental evidence that metals form oxides and sulfides of definite composition. He also recognized that if a metal forms two oxides, each has a definite composition and no product of an intermediate composition exists. With this observation he came close to discovering the law of multiple proportions. In 1808 John Dalton published the first part of the book New System of Chemical Philosophy. He proposed the idea that matter was composed of small discrete particles. Although this concept dated back to
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994
the Greeks of 400 B.C., Dalton's theory was much more far-reaching; it explained the law of conservation of mass as well as the laws of definite composition and multiple proportions. Dalton realized that because compounds are formed by uniting atoms of different elements with different relative weights that can be expressed numerically, the composition of chemical compounds can be expressed quantitatively. He constructed a table of atomic weights, but because of his poor data he was unable to demonstrate the simple relationships that intuitively he knew must exist. Also in 1808 Joseph Louis Gay-Lussac published a paper on the combining volumes of gases. He consistently found that the volume ratios in gaseous reactions were small whole numbers. This result seemed to contradict Dalton's atomic theory because if one volume of CI and one volume of H gave two volumes of HC1, then the "atoms" of CI and H must divide—a logical impossibility if the atomic theory were true. Amedeo Avogadro reconciled the dilemma in 1811 by assuming that equal volumes of gases under the same conditions contained the same number of particles, which he called "molecules." He reasoned that these gaseous molecules split into "half-molecules" when they react. In effect, he supposed that elemental gases contained more than one atom, but he
never used the term atom. Dalton refused to accept Gay-Lussac's law and thus could not appreciate Avogadro's remarkable and revolutionary insight. We know today that Avogadro's reasoning was correct, but the chemists of his day either rejected or simply ignored his hypothesis. Efforts in 1814 by André Marie Ampère and in 1826 by Jean Baptiste Dumas to revive Avogadro's hypothesis went unnoticed. The world of chemistry was not ready for Avogadro. Sadly, it would take another 50 years of confusion about atomic weights before Stanislao Cannizzaro would successfully revive Avogadro's hypothesis. During this period of confusion, practical analysts with little interest in theoretical matters were doing very accurate gravimetric analyses of metals, minerals, and water. Among these were Richard Kerwan, who published an outstanding book listing all the references on water analysis since the time of Bergman; Martin Heinrich Klaproth, who discovered many elements in his analyses of minerals; and Louis Nicolas Vauquelin, who was widely known for his pure reagents. The dominant figure of this period was Jons Jakob Berzelius. During an extraordinarily productive decade between 1807 and 1818, he devised the system of chemical symbols and notations we use today and, through a series of elegant analyses, determined the atomic weight of many of the elements. Unfortunately, he also devised his dualistic theory, which prevented him from accepting Avogadro's hypothesis. Nevertheless, his analytical expertise was nothing short of incredible, and the atomic weights published in his 1828 table approach today's values, if you discount the fact that a few of the elements listed were twice their correct values. Working with the equipment of his day, much of which he improved himself, Berzelius produced a quantity of accurate work that would be a credit to a modern, well-equipped research laboratory. Although Berzelius was one of the world's greatest analysts, it is doubtful that he ever viewed himself as such. In contrast, Karl Remegius Fresenius thought of himself as an analyst from the beginning of his remarkable career. After being an apprentice pharmacist in Frankfurt for several years, he took courses at
the university in Bonn and worked in the as Quantitative Chemical Analysis. private laboratory of Carl Marguart, his Fresenius' small house/laboratory professor of pharmacy. Working mostly would grow to become the worldalone and without instruction, Fresenius renowned Fresenius Institute. By 1855 his taught himself and kept good notes. Mar- laboratory had 60 students, all of whom guart was so impressed with these notes were eligible to receive university credit that he suggested they be published. Anfor the time they spent there. In addition leitung zur qualitativen chemischen Ana- to providing instruction, the institute raplyse, published in 1841, was an immediate idly became known throughout governsuccess and was to see 16 editions under ment, industry, and academia as an analytFresenius' authorship. The 17th edition, ical laboratory. In 1862 Fresenius founded overseen by his son, was translated by Zeitschnftfur analytische Chemie, the first C. A Mitchell in 1921 as Introduction to journal entirely devoted to analytical Qualitative Chemical Analysis. Fresenius chemistry. His influence throughout Euwent to the University of Giessen as a lec- rope and the entire world was enormous. turer, worked in Liebig's laboratory, and obtained his Ph.D. in 1842, using his new History of titrimetry to the book—already in its second edition—as 1850s his thesis. Titrimetry is the determination of an element through the measurement of the weight of a chemical necessary to just complete a definite chemical reaction in a solution containing that element. The weight of the chemical is usually obtained indirectly by measuring the volume of a standard solution of that chemical, although for very accurate work the amount of standard solution needed can be measured by weighing it. Industry's need for rapid methods for determining acids, alkalis, carbonates, and hypochlorites provided the driving force for the development of titrimetry. Early development was confined almost solely to France, where it was crudely practiced in the eighteenth century. Endpoints were determined by the "clear point," the cessation of effervescence, or the use of a few plant-extract indicators. François Antoine Henri Descroizilles devised a method for determining the hyHe was hired in 1845 as professor of chemistry, physics, and technology at the pochlorite strength of bleaching solution used in the textile industry. Hefirstadded Agricultural College at Wiesbaden. Bea measured amount of dilute sulfuric acid cause the college administrators repeatedly refused to supply funds for a chemis- containing an indigo indicator to a graduated cylinder and then slowly added the try laboratory, he borrowed money from hypochlorite solution (whose strength his father, a prosperous attorney, bought was being tested) until the color changed a house, and remodeled it into a private analytical laboratory. He opened the doors from blue to pale green. He then read the volume from the graduations on the side in 1848 withfivestudents and one assistant, Emil Erlenmeyer. Only 30 years old, of the cylinder. Descroizilles later devised a method for determining the alkaline Fresenius was already an experienced strength of potash. He used a graduated analyst. Shortly after assuming his duties buretfilledwith dilute sulfuric acid and at Wiesbaden, he published his second book, Anleitung zur quantitativen chemis- controlled theflowby covering a small air hole in the top with his finger. chen Analyse, which ran to six editions and was translated in 1904 by A I. Cohn Gay-Lussac's contributions to titrime-
Industry's need for rapid methods for determining acids, alkalis, carbonates, and hypochlorites provided the driving force for the development of titrimetry.
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994 227 A
REPOR
υ-
try and the widespread use of his methods helped establish titnmetry as part of clas sical analysis. He improved Descroizilles' method for potash and made several im provements in the hypochlorite method, notably the introduction of the redox indi cator. As an analyst, Gay-Lussac is most fa mous for the silver assay method named after him. The French government, which was losing money because of errors in the fire assay of silver, asked Gay-Lussac in 1829 to devise a rapid and simple method with a relative error of 0.05%. He made a 100-mL pipet that would deliver repeat edly with very high precision, and he pre pared a standard chloride solution that he adjusted so that the delivery of the pipet was equivalent to 1.0000 g of Ag. He then dissolved a bullion sample expected to contain just over 1 g of Ag and added one pipet full of the standard solution. After vigorous agitation, the silver chloride pre cipitate was allowed to flocculate and set tle to the bottom of the flask. Using a 1:10 dilution of the standard chloride solution, he continued the titration in 1-mL incre ments until no more turbidity was pro duced in the supernatant. He achieved a relative accuracy and precision better than the requested 0.05%, and to this day no improvements on this assay method have been made except for the use of a potentiometric endpoint for deeply col ored solutions or alloys of silver contain ing tin. From 1835 to about 1855 many differ ent titrimetric methods were developed but not widely used. As a result of GayLussac's work, however, titrimetry be came known outside of France, especially in Germany and England. It was not yet possible to establish a general system of titrimetry because the concentrations of standard solutions had no chemical basis; there were no unique atomic weights and stoichiometry was not understood. Houton de la Billardière discovered the usefulness of iodine in titrimetry in 1826, and in 1853 Robert Wilhelm Bunsen published an excellent paper on iodimetry describing the determination of more than 20 elements. In that same year, Karl Leonhard Heinrich made a great advance by recommending the use of sodium thiosulfate for titrating iodine. Nevertheless, the famous analysts of the period remained 228 A
contemptuous of titrimetry. Berzelius never used it, and Fresenius recommended that it not be used for important analyses. Karl Friedrich Mohr did much to overcome the difficulties of titrimetry. He studied under the direction of Heinrich Rose and Leopold Gmelin, Bunsen's predecessor at Heidelberg. Mohr took over his father's pharmacy in 1830 and in his spare time experimented with various titrimetric methods. Because he remained outside academia, he was never regarded as a scientist in Germany, but his contributions to titrimetry were many. He introduced the use of potassium chromate as an internal indicator for chloride determination (Mohr method), oxalic acid as a primary standard for alkalimetry, ferrous ammonium sulfate (Mohr's salt) as a standard for oxidizing agents, and the idea of back titration. As for laboratory equipment, he invented the cork borer, the Liebig condenser, the Mohr pinch-cock, the pinch-cock buret, and calibrated pipets. After his book Lehrbuch der chemischanalytischen Titrirmethode was published in two parts in 1855 and 1856, titrimetric analysis became widely known in Europe. The publication of Mohr's books marked the end of the early history of titrimetry, although in 1883 Johan Gustaf C. T. Kjeldahl developed his well-known method for the determination of nitrogen. He discovered that sulfuric acid could be used to digest organic materials, converting free ammonia and "organically bound" nitrogen to ammonium sulfate. After add-
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994
ing excess strong caustic, he distilled the liberated ammonia into a known excess of standard sulfuric acid and determined the unreacted acid by a back titration. The method has remained unchanged, except for the use of various catalysts to aid in the digestion and Lauos Winkler's discovery that the distilled ammonia can be absorbed into a boric acid solution and titrated directly with standard acid. Winkler contributed more than 200 original papers on gravimetry and titrimetry and is probably best remembered for his titrimetric method for the determination of dissolved oxygen in water (published in 1888). The Karlsruhe Congress of 1860 By the late 1850s classical analysis had come about as far as it could without a consistent table of atomic weights. Atomic weight tables varied among different countries and sometimes among different laboratories within the same country. To resolve this problem, Friedrich August Kekulé—with the help of Charles Adolphe Wurtz and Carl Weltzien—invited delegates to an international congress at Karlsruhe, Germany, in 1860. About 140 of the world's leading chemists attended. Cannizzaro, a young professor at the University of Genoa, who for some time had realized the value of Avogadro's hypothesis in resolving the problems surrounding atomic weights, addressed the Karlsruhe Congress with great passion and pedagogical skill. He showed how Avogadro's hypothesis could be used to
ing place in Europe. Graduate schools did establish the molecular weight of a gas not develop in the United States until near and demonstrated that by comparing the vapor densities of a series of gaseous com- the end of the century. From about the pounds of a particular element, the molec- middle of the century, Germany was the place to go for graduate education in ular weight and atomic weight of that elechemistry. Justus Liebig, who established ment could be determined. Although a teaching laboratory at the University of Cannizzaro pleaded with his colleagues Giessen, generated remarkable enthusifor the adoption of atomic weights based asm and camaraderie among his students, on Avogadro's hypothesis, he was unable and his laboratory became a model for to sway congress members in his favor. other graduate programs and teaching The congress adjourned, having reached no agreement on atomic weights, laboratories throughout Germany. but it was in adjourning that the key event of the congress occurred. As the delegates departed, Cannizzaro's colleague, Angelo Pavesi from the University of Pavia, passed out reprints of Cannizzaro's 1858 publication, Sunto di un corso di Filosofia chimica (Sketch of a Course of Chemical Philosophy). This paper set forth clearly what Cannizzaro had been teaching his students at Genoa. Although his paper had gone largely unnoticed in the literature, it did not go unnoticed in the hands of the delegates. Julius Lothar Meyer read it twice on his homeward journey. Everything became clear to him, as if scales fell from his eyes. Dmitrii Ivanovitch Mendeleev later said that even though no agreement was reached by the congress, the truth of the law of Avogadro as advocated by Cannizzaro soon convinced everyone. Without the insight that both Meyer and Mendeleev gained from Cannizzaro's paper, it is doubtful that they independently would have gone on to work out the periodic William Francis Hillebrand was born in table in 1869. Honolulu seven years before the KarlsIt is hard to overestimate the impact of ruhe Congress, and his story is perhaps the best example of how classical analysis, the events surrounding the Karlsruhe particularly gravimetry, was brought to Congress. The confusion over atomic weights almost completely disappeared in the United States. His father, a Germanjust a few years, and analysts were able to born medical doctor, established a practice in Hawaii because its climate would write down correct stoichiometric formuhelp him to recover from a serious illness. las for their precipitates. Gravimetry was Because his father intended to return to placed on solid, although still somewhat Germany with the entire family, Hilleempirical, ground. Titrimetry also benebrand was sent to Cornell in 1870 to prefited from this great advance, but it (and, pare for study at a German university. to a lesser extent, gravimetry) lacked a firm scientific basis until the development The family moved to Bonn in the sumof physical chemistry at the end of the mer of 1872, and Hillebrand had to choose nineteenth century. a career. He had no interest in medicine and felt he lacked the mental qualifications for law or engineering. His father Gravimetry comes to the suggested chemistry. Remembering with United States pleasure his study of the basics of chemisIn the mid-nineteenth century almost all new developments in chemistry were tak- try back in Honolulu, Hillebrand decided
The significance of Jiittebrand's worf^iay in the perfection of a separation schemefor matenais as compteras carbonate ana silicate rocks.
to give chemistry a try. He entered the University of Heidelberg in 1872 and studied under the guidance of Bunsen and Gustav Kirchhoff, earning his Ph.D. in 1875. He stayed on at Heidelberg for another year of research and then spent three semesters with Rudolph Fittig at Strasbourg. Realizing that organic chemistry did not appeal to him, he decided to finish his studies at the mining academy at Freiberg to supplement the training in mineral analysis, which he received under the direction of Bunsen. He had made up his mind to become an analytical chemist. He returned to the United States in 1878 and, failing to find immediate work in the East, made his way in 1879 to Leadville, CO, where he became the third partner in a small assaying firm. Samuel F. Emmons, an occasional customer who was in charge of the Rocky Mountain Division of the newly formed United States Geological Survey (USGS), offered Hillebrand a job as a chemist. Hillebrand considered Emmons' offer the opportunity of a lifetime and quickly accepted. He remained in the Denver laboratory of the USGS until 1885, when he was transferred to the Washington USGS laboratory to work under the direction of the chief chemist, Frank W. Clarke. Hillebrand's careful work set a new standard of excellence in the analysis of rocks and ores. During his 29 years at the USGS, he made more than 400 complete analyses of silicate rocks. The significance of his work lay in the perfection of a separation scheme for materials as complex as carbonate and silicate rocks. Even in the late nineteenth century, the problem wasn't so much the lack of a suitable method for the final determination of an element as it was the lack of suitable separation methods preceding thefinaldetermination—a problem that still exists. In 1897 Hillebrand wrote a 50-page introduction to USGS Bulletin No. 148 on the methods of analysis of silicate rocks, and this was quickly translated into German. In 1900 the introduction was revised, enlarged, and printed as an independent document, Bulletin No. 176. The next edition appeared in 1907 as Bulletin No. 305. It included carbonate rocks and was also translated into German. The series culminated in 1919 with the well-known Bulletin No. 700, a book of 285 pages.
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994 229 A
REPORT
Hillebrand was unquestionably one of the world's leading analysts when he was called to be chief chemist at the National Bureau of Standards (NBS) in 1909. Administrative duties in his new position limited his time at the bench, but the steady stream of books and papers continued. He greatly contributed to the increase of the standard samples program at NBS, expanding it from a few cast irons to more than 60 materials. (Today, the Standard Reference Materials Program at the National Institute of Standards and Technology [NIST, formerly NBS] has more than 1200 standard reference materials available, covering a broad range of materials that are certified for chemical composition and/or physical property.) Hillebrand had a talent for gathering around him capable people, and his success in bringing Gustaf E. F. Lundell from Cornell in 1917 is a good example. In 1923 he and Lundell began co-authorship of Applied Inorganic Analysis. Hillebrand died in 1925 before the work was completed, but by 1929 Lundell had finished the book, which became known as the "analyst's bible." Together with James I. Hoffman, Lundell wrote a companion volume in 1938, Outlines ofMethods of Chemical Analysis, which became another classic. (Hoffman and Harry A. Bright brought out a second edition of Applied Inorganic Analysis in 1953.) Lundell was appointed chief chemist at NBS in 1937, and under his leadership the 230 A
scope and renown of its efforts in chemical analysis grew. He had a talent for picking the right person for a particular assignment, and he had an almost uncanny ability to sense whether work on a project that had hit a snag should be stopped or continued. Often his timely encouragement resulted in the solution of a seemingly hopeless problem. In 1933 Lundell published his well-known paper, "The Chemical Analysis of Things as They Are" (2). The continuing relevance of his insights is remarkable, and this paper should be required reading for all analytical chemists. Physical chemistry applied to classical analysis Once the confusion over atomic weights was settled, the stage was set for the maturation of gravimetry. Although many refinements would result from the application of physical chemistry to gravimetry, putting it to practical use did not require an understanding of the rate of precipitate formation, growth of crystalline precipitates, adsorption and occlusion of impurities, and the aging of precipitates. Such was not the case with titrimetry. Correct atomic weights certainly were as necessary for titrimetry as for gravimetry. However, input from the newly emerging physical chemistry would be essential for titrimetry to reach maturity. Physical chemistry did not emerge until the last third of the nineteenth cen-
Analytical Chemistry, Vol. 66, No. 4, February 15, 1994
tury. A great breakthrough was made between 1864 and 1879 when the Norwegian brothers-in-law Cato M. Guldberg and Peter Waage formulated the law of mass action. In 1884 Jacobus Henricus van't Hoff made an elegant derivation of the law of mass action based on thermodynamics. His clear and inspiring treatment of chemical dynamics brought the entire subject of reaction kinetics and equilibria before the chemical world. Certain anomalies that occurred when the laws of physical chemistry were applied to solutions of electrolytes were explained in 1887 when the revolutionary theory of electrolytic dissociation was published by Svante August Arrhenius. It was Wilhelm Ostwald who recognized the importance of the work done by van't Hoff and Arrhenius. Through his books, research, and personal contacts, Ostwald was influential in spreading the ideas of the new physical chemistry. With van't Hoff he founded the journal Zeitschriftfurphysikalische Chemie in 1887, and he championed the cause of physical chemistry as a science in its own right In essence, he "organized" physical chemistry at the end of the nineteenth century. In 1894 Ostwald published Die wissenschafllichen Grundlagen der analytischen Chemie (The Scientific Foundations of Analytical Chemistry) and thereby began to put classical analysis on a scientific basis. In the preface he noted that although the technique of chemical analysis stood at a very high level, its scientific treatment was almost completely neglected. Ostwald's book was a significant start toward correcting this deficiency. He discussed precipitation in detail, including the increase in particle size of standing crystalline precipitates—a process that became known as "Ostwald ripening." The most important part of the book dealt with chemical separations. By combining the law of mass action and Arrhenius' theory of electrolytic dissociation, he introduced the concepts of dissociation constants and solubility product constants. Although Ostwald's little book broke new ground and was soon recognized as a classic, it suffered some significant omissions. For example, the phase rule of Josiah Willard Gibbs was not mentioned. Perhaps Ostwald was simply unaware of Gibbs' work because it had been pub-
lished in a rather obscure journal. However, it is difficult to understand why he made no mention of the work of Walter Nernst, who originated the famous Nernst equation while working in Ostwald's own laboratory in 1889. Nernst explored many analytical applications of his equation, and he is rightly considered the father of modern electroanalytical chemistry. Titrimetry comes to the United States As was the case with gravimetry, titrimetry developed in Europe and then came to the United States, chiefly through the efforts of Izaak Maurits Kolthoff, who came to the University of Utrecht in Holland in 1911. Kolthoff's lack of certain prerequisites demanded by the chemistry department was a blessing in disguise because it brought him under the tutelage of Nicolaas Schoorl in the school of pharmacy. Schoorl had studied under the direction of van't Hoff and was aware of the recent advances in physical chemistry. He immediately recognized Kolthoff s talent and encouraged him to carry out independent research. Kolthoff had trouble understanding the proper selection of indicators for acidbase titrations and therefore launched his own investigation. He had already acquired a used copy of Ostwald's 1894 classic, and he was further encouraged by the 1909 paper of the Danish physiological chemist S.P.L Sjirensen, who introduced the concept of pH. Kolthoff was further inspired in 1913 by the work of Joel Hildebrand, who used the hydrogen reference electrode in electrometric titrations. Kolthoff borrowed pH measuring equipment, but within a year he had devised his own potentiometric apparatus. In 1915 he published his first paper on the titration of phosphoric acid. Kolthoff quickly realized the importance of physical chemistry to analytical chemistry, but in those early years, sometimes the going was rough. Not recognizing the significance of his early work, chemistry department faculty members sometimes criticized him. Some went so far as to try to block him from publishing and lecturing. Despite such criticism, Kolthoff forged ahead and in 1918 presented his Ph.D. thesis on the "Fundamentals of Iodimetry," a topic he returned
to again and again over the years. In 1914 the Danish physical chemist Niels J. Bjerrum published a book showing how to calculate both the shape of neutralization curves and the titration errors in visual endpoint determinations. Bjerrum's work inspired Kolthoff to develop a theoretical interpretation of all the methods of titrimetry. (This work resulted in his famous two-volume book, first published in German in 1926 and translated in 1928 and 1929 as Volumetric Analysis by N. Howell Furman of Princeton University. A three-volume revision of the work appeared between 1942 and 1957.)
