Chemical origins of color - Journal of Chemical Education (ACS

---

Mary Virginia Orna, O.S.U. College of New Rochelle New Rochelle. New York 10801

I

The Chemical Origins of Color

It was not until Newton's experiments in the 17th century that a firm theoretical founda tion regarding the nature of color was laid.

Color is a property of materials that has been an integral part of human experience in every age and civilization. It has caused man to wonder about its origin (I) and experiment in its production. Historically, the use of color was chiefly an art which developed slowly into an organized hody of knowledge so that by the Fifth Century, B.C., the Greeks were writing treatises on color harmony, perspective, and the preparation of niements. Thev also succeeded in exnandine the artist's paiette to include-white lead, red lead, &d veriilion. I t was left to the practical, aggressive Roman businessman to commercialize color usage by the manufacture and distribution of "mass-produced" colored items. This was the beginning of a more advanced, hut strictly empirical, color technology, and it was not until Newton's experiments in the Seventeenth Century that a firm theoretical foundation regarding the nature of color was laid ( 2 ) . Today, color science plays a major role in business, science and industry. I t is one of the few disciplines that cuts across the boundaries of art, biology, physics, psychology, chemistry, geology, mineralogy, and many other fields. There is hardly an object or a substance in nature that is not colored, and virtually every commercially marketed item today is either deliberately colored or de-colored (3). Chemists have always had a great interest in color. As early as 1909. N. Bierrum remarked that one of the most invariant properties of a given chemical species is its color, that is, its absorption spectrum, and he attributed the colors observed largely to the ligands in the first coordination sphere of metal complexes (4).Since that time, many a student in introductnry chemistry has monitored a chemical change by observing a color change. Virtually every quantitative analysis laboratory manual includes analysis by permanganate and dichromate redox titrations; the color changes of bromcresol green, methyl red, and crystal violet are universally used to detect the endpoints of particular t w e s of titrations: EDTA com~lexeswith certain metals yield intense blues and reds; the fdrmation of a deep hlue precipitate of Fe3[Fe(CN)& upon reaction with halide is used to determine the place of the Fe2+-Fe" couple in a potential series of the halogens, and many a freshman has seen the dramatic transformation of pale blue aqueous copper(I1) sulfate to a deep royal hlue upon the addition of ammonia. Colored compounds are all around us and we do not hesitate to utilize their properties for specific purposes. But how often do we bother to classify the compounds of color for our students or attempt to explain the nature of color? In my own experience, we often refer to the fact of color, but only rarely and sketchily do we look into the fundamental reasons for its occurrence. I would like to suggest that it may he very worthwhile to pay more attention to color and its chemistry than we have in the past. With the movement toward more descriptive chemistry in chemical education, color provides a perfect link between an easily observed and described property and an underlying theory. With the rise of s~ectroscoov .. in the undereraduate curriculum, visible spectroscopy provides a very familiar 478 1 Journal of Chemlcal Education

starting place and frame of reference within which one can introduce the other spectroscopies. Color lends itself to the inclusion of interesting, enriching, and exciting topics in introductory and advanced courses. I t could even he the pivot for a course for nonscience majors, or a c o m e for intermediate or advanced chemistry majors, or part of a special topics course or seminar. It is a topic that has stimulated the imaginations of artists and poets over the centuries; perhaps it can stimulate chemical imaginations to a greater degree than it has in the oast. Finallv. i t is a verv-nractical tonic with manv . applications to indus&ial problems and needs. Color usage extends to l i e h t i n ~oroblems in work areas. streets. airstrios. theatres andplac& of public assembly; it'is a fundamenial part of the dye, pigments, printing, textile, plastics, photographic, and entertainment industries; color coding of lights, sims, tools, electrical wirine and utilities conduits are Dart of o;r everyday lives. Color thus provides a most appropriate interface between academia and industry. I t is the purpose of this paper to discuss the topic of cnlor science with emphasis on the main classes of compounds which exhibit color in order to provide students with a basic introduction to color science and chemical educators with some ideas to include in their respective chemical curricula. The Nature of Color Color is fundamentallv a suhiective phenomenon. I t is the result of a stimulus received hy the eye and interpreted by the brain, and no one can explain the nroduction of color without taking three factors intdaccount:the light source, the object i t illuminates, and the eye and brain which receive and perceive the color. 1) The Light Source. Electromagnetic radiation has been characterized quite elegantly by the Einstein-Planck relationship, E = hv, where E is the energy of the individual photon, h is the Planck constant and u is the frequency of the radiation in sec-'. A convenient value for h is 4.136 X lo-'" eV-sec so that all energies may he rumputed directly in elertron volts. Every source of illumination emitb photons of a ranee of enereies. The intensitv of the radiation mav varv ~~"with wavelength to yield a spectral power distribution curve such as that shown in Fieure l a for a tunesten source. Since the distribution of radiant power varies from source to source, or from time to time in the same source, i t is important to specify

-

-

-

-

Flgure 1 a. Specwal power dlsw~butloncurve-mcandescent light b. Reflsctance cuve of m o d l t y q oblecr, c, homulus facolor cuve (oofen pacewed as red).

Recent research indicates that three-color information is somehow processed in the retina and encoded in two-color, on-offsignals which find their way to higher visual centers. the tvne .. of illumination under which an ohiect is viewed. A sour(.e which emiw energy continuously over the limited resnmse ranee of the human eve. from about 380 to 720 nm, and with appreciable intensities&all wavelengths is perceived by the eve as white and is therefore descrihed as "white" light. On the other hand, a sodium arc lamp exhibits a spectrum of discrete lines with its most intense line a t 590 nm and is perceived as "yellow." Dispersion of white light with a prism or grating yields the familiar spectral colors, ranging from red a t around 700 nm to violet a t around 400 nm, a t various viewing positions. 2) The Object. Every illuminated object modifies the light which falls upon it in several of the followingways: reflection, transmission, absorption, scattering, dispersion, interference, and diffraction. Such modification gives rise to what we perceive as black, white, and colored ohjects which may he transparent, translucent, or opaque. Although light modification by an ohject is a very complex phenomenon, the process of greatest interest to the chemist is selective absorption which vields a characteristic transmission or reflectance curve for an absorbing species. I t is this curve superimposed on the spectral power distribution curve of Figure l a which provides the stimulus curve for the human eye or other suitahle detector as illustrated in Figure l b (5). 3) The Detector. A variety of color detectors is available for study, and perhaps the most familiar is the human eye itself, together with its response areas in the nervous system and the brain. Although this detector has been the subject of much study, it is still not clear how it works, hut all other visible light detector systems devised by man have tried to duplicate its results in one wav or another. The most imiortant structure in the human eye for the nercention . of color is the retina. which contains the cone cells responsible for color vision (6).One of the earliest theories of color vision was that of Thomas Young (1802) elaborated upon by H. von Helmholtz around 1852. The combined YoungHelmholtz theory postulates that since the retina responds in a t least three different ways to different colors, there must he three different kinds of r e c e ~ t o neach , of which is sensitive to a particular portion of the visible spectrum (7).It took more than a century to obtain the cone-specific spectrophotometric data that showed that there are indeed three types of cones, each of which contains one of three light-sensitive pigments (a,!?).The existence of this three-color, three-receptor physiological system is consistent with the three-primary postulate of Young and Helmholtz. However, the fact that virtually every color test given has induced subjects t o name four instead of three unique primaries gave rise to the opponentprocess theory of Hering (lo), which holds that yellow must he counted as a primary color along with red, green, and blue. This theory takes into account the complementarity of redgreen and yellow-blue and assumes that these four colors, together with black and white, form three pairs of unique sensory qualities which are mutually exclusive or "opponent" to one another. Recent research indicates that three-color information is somehow nrocessed in the retina and encoded into two-color, on-off signals which eventually find their way to the hieher visual centers via a nathwav about which we still know very little (11). I t is nossihle. without much knowledee of the mechanism of cdo;sensiti&y, to measure the relatiGe responsivity of the eve to various waveleneths of visible light. The result is a spectral response curve, and a complete description of the color stimulating the eye would then involve a combination

.

of this response curve with the stimulus curve of Figure lb. If the stimulus contains all the wavelengths of the visible spectrum, the eye perceives white light, hut when only some of the wavelengths are nresent. the eve ~erceivescolor. For example, if the &en wavelengths are a6&rhed hy a modifying ohiect. the eve sees the remaining wavelenrrths, that is, the hlie-rkd combination we call magenta, t h e complement to green. Color Modification I t is obvious from this discussion that the only color-producine factor over which the chemist has anv control is the modif;ing ohject. Although most colorimetric measurements are made with reference to a articular standard illuminant and are described by a built-in "standard observer" response curve (specified by the International Commission on Illumination in 1931), the chemist ultimately is responsible only for the reflectance or transmittance curve of the modifying object since, once it leaves his hands, i t may he viewed under any light source and by any observer or detector. T o be sure, the chemist may he required to design a modifying object for observation under specified illumination conditions. Chemists mav wish to modifv colors or observe colormodifying comp&nds for a variet; of reasons. On the practical side, color matching and color formulation are very important in many different industries. From a more theoretical standpoint, spectroscopists are interested in all the electronic processes a material can undergo, and these extend throughout the ultraviolet as well as the visible rezion of the swctrum. Spectrascopic data can yield a great deal of information ahout how structural changes affect the energy spacings in molecules, and once a theoretical framework for color modification of molecules is laid, the wheel comes full circle when the industrial chemist utilizes theory to obtain the results he wants. However, the mast valid reason of all is still scientific curiwity: chemists are basically people who want to know why. In order to examine the reasons for the existence of color in material ohjects, we must constantly refer to the principles of quantum mechanics which state that only certain discrete energy levels are permitted for electrons in bound states. As a consequence, there are well-defined energy differences between these allowed levels, and when electromagnetic radiation interacts with an ohject, only those wavelengths whose energies correspond exactly to the energy level differences in the ohject will be absorbed. The absorbed energy is utilized to excite a soecies from one electronic level to another. and it is these electronic transitionsthat give rise torolor and color modification in atoms, molecules and crystals. If, in the absence of any elertromagnetic radiation, all the electrons in a molecule are in their lowest availahle enerev levels, the molecule is said to be in its ground state. ~ h s o r ~ t i of o nsuitable enereies in the form of electromagnetic radiation can promote the species to excited states. All atoms, molecules andcrystals whether colored or colorless, exhibit this hehavior since there is no difference in principle between electronic transitions giving rise to absorption in the visible and ultraviolet regions. The appearance of color, caused hy absorption over a restricted wavelenrrth range, is determined by the sensitivity of the eye. But color is not connected with anyone special feature of molecular structure (12). If the energy spacings in a molecule are large, that is, greater than ahout 3.2 eV, photons in the ultraviolet are required to excite i t from the &ound state t o an excited state. But. if the molecule can be modified structurally so that its dnergy ~

~

~~~~

~~

Volume 55, Number 8, August 1978 / 479

-

Accordina r * transitions and r - to the MO model, n transitions may occur in the visible region. spacings decrease, we will observe a shift in the absorption maximum to longer wavelengths corresponding to the lower energies required for electronic excitation. This is called a bathochromic shift. It is a necessary, hut not sufficient, condition for those materials which exhihit color to have sets of electronic energy levels separated by no less than about 1.7 eV (720 nm) and no more than 3.2 eV (380 nm). and thus ahsorl; in the "isible region Because vibiational A d rotational enerev levels are suoerimoosed on the electronic levels in a m o l e h e , photons at a number of wavelengths on both sides of the ~rincioalahsorotion band are alsoabsorbed. eivine rise t o theabroad bands iharacteristic of ultraviolet and visible soectra. The major electronic processes in materials which give rise to color t ~ yselective absorption mav he classified as follows 1) oreanic conmounds ~. transitions in coniuested ,2) intern,olrrular charge transfer transitions 3) intrarndeednr c h a r g ~trnmfrr transltionr 4) crystal field transitions 5) band transitions

Transltlons in Con/ugaied Organic Compounds In 1876,O. N. Witt (13)proposed that in order for an organic compound to exhihit color, it must contain an nnsaturated group called a chromophore. Some common chromo-C=C-, A==, 4 s H s . However, phores are -N=N-, their presence does not mean that a molecule will necessarily possess a color. Witt also proposed that the presence of other groups, called auxochromes, such as -OH, -NH2, and -NHR, served to strengthen and deepen the color of a molecule. A molecule containing a chromophore but not an auxochrome is called a chromoeen. Addition of auxochromes or accumulation of chromophores can lead to the development of color in a chromoeen. The basis for Witt's emoirical observations and thosiof others can be seen in three models formulated using the principles of quantum mechanics and described below. The Molecular Orbital Model The three types of valence electrons giving rise to electronic transitions in organic molecules are those involved in single bond ( a )and double bond (u) formation, and non-bonded (n) electrons associated with heteroatoms. When the constituent atoms of a molecule are a t the equilibrium distances characteristic of the stable molecule, the atomic orbitals can he linearly combined to form molerular orhitals. The total wave function of the molecule is taken an a romhination of these molecular orhitals. Those rnolerular nrhital4 where the dectron density is gvatrst hetween the nuclei are termed hondina orbitals and have energies lower than the contrihutingatomic orbitals. Those molecular orhitals with very small electron densities between the nuclei are termed antibonding orbitals, designated by a "*." Saturated molecules. which can onlv undereo a a* transitions, require large excitation energies and atworb in the far ultraviolet. Nun-honded electrons a r t less riehtlv held and thus n n* transitions require far less energy for molecular excitation and eenerallv occur in the visible reeion. Since n-orbital overlap is not as great as in a-bonding, u a* transition energies take on values intermediate to the other two twerr and occur in the near ultraviolet and visihle regions. Extended r-coniugation leads ton bathochrwnicshift in rhe n +* absorptibnmaxima. For example, ethene contains the -M- chromophore but is colorless because there is a 7.52

-

-

-

480 1 Journal of Chemical Education

r*

eV energy gap between its most closely spaced energy levels and i t absorbs in the far ultraviolet. Addition of more -C=Cchromophores leads to energy levels tbat are more closely spaced. Ultimately, as in the case of B-carotene, containing 11 such chromophores (see formula below), the energy gap is only around 2.5 eV, and an absorption maximum occurs in the blue region of the visible spectrum.

An example of the effect of the presence of heteroatoms can be seen in comparing stilbene, a colorless compound, with ambenzene, which is orange (3.14.15).

Molecular orbital theory can predict the intensities of absorption hands for simple molecules.

We will briefly examine the nature of each of these types of transitions and look a t some examples.

-

-

-

The Valence Bond Model

This model is very familiar to chemists in its semi-intuitive qualitative extension known as resonance theory, where the hathochromic shift we obsewed above as aresult of extended conjugation is treated in terms of an increasing number of contrihutine structures with similar relat,ive stabilities. Valence hond 'theory is very concerned with the contributions made t o snectra bv ionic states and can also oredict the expected lockion of"absorption hands.

The Free Electron Madel This is a modification of the MO theory which singles out one or several MO's for semi-quantitative treatment. This model is particularly applicable t o conjugated u-electron systems, and since most dyes and organir pigmentn fall into this category, it is quite an appropriate method. This model assumes that the conjugated r-electrons are in a well of constant notential ener& whose houndaries are rouehlv a little longe; than the leng% of the carhon chain, and tgat ihe electrons are free to move within the system. These simplifying assumptions reduce the problem t o solving the Schrodinger equation for a particle in a one-dimensional box, and this leads t o the calculation of the theoretical absorption wavelength by A = 8mcL2/(n 1)h. where m is the mass of the electron. c is the veloci&of light; L is the length of the potential energy well, n is the number of conjugated T-electrons and h is the Planck constant. This model has yielded calculated values very close to the observed values for several series of dyes

+

(16).

In(ermolecular Charge Transfer Transitions There is a whole group of organic compounds for which the intramolecular model discussed above must be modified. For examole. a 1:l mixture of auinone and hvdroauinone in an . . alcohblic solution yields bdautiful dark green crystals of a material tbat could onlv be formed from one molecule of quinone reacting with one molecule of hydroquinone to form a weakly bound complex somewhat like the following

<

-

-

The inter-moiety bonding in such a complex is not conjugated,

and the species is not really a single molecule; yet i t absorbs both violet and red light and thus appears green. T o explain this phenomenon, R. S. Mulliken (17,18) has sueeested that ahsorotion of a nhoton hv such a comolex can result in a charge rearrangement involving electron transfer from one component of the complex to the other. According t o Mulliken's theory, one component of the complex may he considered the electron donor, D,and the other the electron acceptor, A. I t is now possible to think of the complex as existine as a hvhrid between two extreme forms. the no-bond form: (D, ~j and the structure (D+-A-) in which the two soecies are hound together hv a covalent chemical bond. In tile no-bond form, each of the species retains its own electrons. The complex is formed when the donor species absorbs radiant energy and its electron enters the next highest energy level available, that is, the lowest lying vacant molecular orhital of the acceptor species. We may think of this phenomenon as one involving an intermolecular transition and the no-bond and bonded forms as the resonance structures contributing to intermolecular resonance. In the quinone-hydroquinone example, quinone is the donor and the hydroquinone is the acceptor. There are two types of donor compounds, those that have either r-electrons or electrons in non-bonding orbitals available. The former are called n-donors and include molecules containing the henzene ring, double or triple bonds; the latter type compound usually rontains an element with lone pairs of electrons such as nitrogen or oxygen. There are many different kinds of acceptors, both organic and inorganic. When certain atoms or groups of atoms such as -NO2 and -C1 are oresent in a comnound whirh contains hvdroeens on nearbv carbon or oxygen atoms, for example, in picric acid or chloroform. a hvdroeen bonded comnlex becomes oossible. For example, chloroform forms a colorless charge transfer complex with acetone that looks like this

--

-

HC

7' '

\ C=O ..H-C-CI / H ,C n I

and picric acid forms numerous bright yellow solid picrates with a host of electron donors. There are many inorganic acceptors as well. The halogens are a good example. Iodine dissolved in CCl4 exhibits its normal violet color, which means that i t is absorbing yellow-green light. However, solutions of iodine in henzene are deen red to reddish-brown, indicatingformation of a r-complexwith benzene (19,220). Intramolecular Charge Transfer Transitions The charge transfer phenomenon is not limited to organic molecules nor is it limited to the intermolecular form. I t was first recoenized hv Przibram 121) . . in 1923 that ordinarv tahle salt in the gas phase will exhibit photon absorption a t around 234 nm. For our purposes we will consider that, in the gas phase, sodium chloride consists of a single sodium ion hound t o a single chloride ion. The observed ultraviolet absorption hand at 234 nm may he attrihuted t o a transfer of an electron from C1- to Na+ to form a covalently hound NaCl molecule. Many other salts related to sodium chloride exhibit charge transfer suectra in the ultraviolet or far ultraviolet rezions and are examples of intramolecular transitions in organic compounds (15,22). Although NaCI was a familiar example to use to illustrate the charge transfer phenomenon in inorganic compounds, NaCl is colorless because the absorption of radiant energy takes place in the ultraviolet region. In fact, most inorganic compounds exhibit charge transfer spectra in the ultraviolet, hut those that absorb in the visible region have intense and very dramatic colors. For example, when Few is placed in an aqueous solution containing CI-, i t has the possibility of associating with C1- to form the species containing one to six chlorides, such as Fe(H20)&12+, etc. All of these species ahsorb intensely in the ultraviolet and the bands extend into the

-

blue region of the visible spectrum, thus accounting for the vellow to oranee color of FeCL solutions. These color-~roducing absor&on hands have ieen attributed to the charge transfer nhenomenon. noth her iron complex that is colored is that of iron(1II) with thiocyanate. The CNS- tends to transfer its charge to Fe3+ to yield an excited complex partially formulated as Fez+-CNS, which absorbs intensely around 500 nm and is deep red (23, 24). Other examples are some mercury, bismuth, and lead compounds that exhibit brilliant colors. Also, bright yellow chromate5 and deep violet permanganates can he formed. If we consider as a ligand any molecule or ion hound to a metal, then we can generalize the charge transfer process by stating that i t occurs when an electron is tranferred from an orbital lying principally on the ligand to an orbital lying principally on the metal, or vice versa. The former case is known as ligand-to-metal, or an L M, transition, and the inverse is an M - L process. The process may occur in transition metal complexes as well as others. As a matter of fact, all polyatomic molecules exhihir charge transfer spectra, hut only thwe with ahsorption occurring in the visihle region will he colored. Most charge transfer processes require relatively high energies and therefore usually lie in the ultraviolet or far ultraviolet regions of the spectrum. However, if the metal is L easily oxidizahle and its ligand is easily reducible, M charee transfer transitions mav occur in the visible reeion. Bv t h e s k e token, if the ligand iskasi~yoxidizahle and meG readilv reducible. then L M transitions mav take olace in the viiihle as well'(2.1). Most charge tranqfer trkwtions in the v~sihlereeion - are exhihlted hv comoounds of the d-block elements. There is a very definite periodic trend in the energies required for charge transfer transitions. The cause is that for a given metal, ligand-to-metal charge transfer energies decrease as the ligand becomes more oxidizahle, while metal-to-ligand c h m e transfer energies decrease as the lieand becomes more readily reducible. F& a given ligand, me&-to-ligand charge transfer bands decrease in energy as the oxidation state of the metal decreases, or in general, as the metal becomes more readily oxidizahle. J6rgensen (24,25) has attempted t o correlate these regularities by suggesting that charge transfer transition energies can he related to the eledroneeativitv difference hetween the donor and acceptor orhit& in H complex, and he has called this difference the "optical electronegativity." A special case of charge transfer, namely between two atoms of the same element in two different oxidation states, is also responsible for some very familiar colored compounds. For example, Prussian blue, with the formula KFe(III)Fe(II)(CN)e, has a deep blue color which has been attrihuted to intervalence charge transfer from Fe(I1) to Fe(II1). Although the mixed valence compounds of iron were the earliest synthesized and best-known, numerous other mixed valence pairs have been the subject of extensive investigations (26).

a

-

.

-

-

tee

Crystal Fleid Transltlons The compounds of the d-block metals, taken as a group, exhibit almost every conceivable color, and an ion of the same metal can exhibit different colors in different chemical environments. While charge transfer transitions may he partly responsihle for these colors, another absorption mechanism involving d-electrons is also involved. The d-orbitals of a gaseous metal ion have preferred orientations in space. Graphical representations of these orbital shapes may be found in most modem textbooks of general or inorganic chemistry. In the gaseous free ion, the d-electrons are subjected only to an interelectronic repulsion force which causes a solittine of the d-enerev level into a seouence of levels. ~ h i the n .on is placed in &tion or in a crysh lattice, the surrounding electrir field exerted hv the nearest-neiehhnr ligands causes further splitting of thefree ion levels because Volume 55, Number 8, August 1978 1 481

of the different geometries of the d-orbitals relative to the approaching ligands. The kind of splittine depends upon the number of d-electrons and the symmetry of the field The simplest example is that of Tiq+ in a six-coordinate octahedral field. The single d-electron of Ti3+ experiences no interelectronic repulsion, hut the d,z and d,2-,2 orhitals oriented along the metal-ligand axes are destahilized (have higher energies) while the d, d,,, and d,, orbitals lying along the bisectors of the angles hetween the axes are st,abilized (have lower energies) wit,h respect to the ion in a spherirally symmetrical field. The end r e s d t of the imposition of t,he ligand field is a splitting of the pound state of the metalion into several different levels. the numher of which denends unon the nature of the field and the number of d-elec&ons. since, in this case, the d-orhitals are no longer degenerate, electrons may absorb radiation and he excited from a stabilized d orbital to a destabilized one. These transitions are often referred to as crystal field, or d-d, transitions, and the energies required often correspond to the waveleneths of visible lieht. ~ h k s etransitions cannot occur if the metal ion has n o d l electrons, as in the case of Sc3+. or if the d-orbitals are completely fillrd, n s in the rase of Zn2+.Most of the intermediate d-l~lockelements of the fourth period exhibit rolorsdue tu d-d transitions, and the colors of most minerals and gems can be explained using this model together with that for charge . transfer (27). Charge transfer spectra are usually quite intense and often mask the relativelv weak crvstal field suectral bands. The bright familiar colors of the chlorides, bromides, and oxides of the d-hlock elements are due to both types of transitions, *+ whereas the very pale hlue Cu(OH)2 and the C U ( H ~ O ) ~ion are due to d-d transitions only (28). The pale pink colors of muscovite minerals have been attrihuted to d-d transitions of Fe(II1) impurities suhstituted for Al and Si in tetrahedral sites (29), and the green color of emerald is due to Cr3+ in octahedral Al" sites (30). 'l'he widely variant intensities of charge transfer and crystal field trunsiti~nsaredue tuthe fact that not all transitions take place with the same prohahility. For example, a transition requiring a net change in electron spin will take place much less readily than one in which the electron spin is the same in the initial and final states. Such transitions are called "spinforbidden'' and "spin-allowed," respectively. Forbidden transitions often occur, hut. with low intensities, thus accounting for the very pale colors ohserved in many d-block complexes. The types of forbidden and allowed transitions have been summarized by a set of selection rules which give the conditions for the probability of a particular transition. An additional constraint upnn elertronic transitions is the fact that for the electron makinp the quantum jump, the orbital aneular momentum must chanee . hv. one unit of h/2a. This selection rule is generally expressed as A1 = f1, and is called the Laporte Rule after its discoverer. A consequence of this rule is that., although the s-p, p-d or d-f transitions are allowed, d-d transitions are forbidden since all the d orbitals have the same 1 value and A1 = 0 for a d-d transition. On the other hand, most charge transfer transitions are Laporteallowed and are thus much more intense (probable) than d-d transitions taking plare in the same complex, and thus are likely t o mask them. Several general articles on the crystal field and ligand field theories have appeared in this Journal (31).

cules will cause a change in the energy necessary to promote a species to an excited state, and thus a change in the wavelength of electromagnetic radiation absorhed. In several instances, we have also seen that such energy level modifications ran he accomplished by allowing ligands to approach isolated at.oms or to allow a group of isolated atoms to approach one another. In the case of the formation of crystals or very large mdecules. atoms with onlv definite allowed enereies annroach one anothkr and interact kith one another in such a way as to multiply the number of allowed energy states. These states are so rlosely spaced that a collection of such states is often referred to as a "hand," and the spacing between hands, that is, the energy forhidden to the crystal's electrons, is called the band gap. In metallic crystals, the valence electrons are very loosely held because it is the upper energy levels that are the most susceptible to band broadening. This allows metallic crystals to absorb electromagnetic radiation of almost all wavelengths hecause the absorbed radiation onlv serves to impart kinetic energy to the electron. on-metalfie crystals, however, contain very few free electrons and a photon will he absorbed only if it has enough energy to overcome the hand gap and break an electron free from a bonding orbital. These crystals are usually transparent in the infrared because infrared radiation is not energetic enough to free the electrons and so infrared radiation is not absorhed. However. the crystals are usually opaque in the ultraviolet because "ltraviolet lieht is enereetic enoueh to shake louse the electrons and thus is absorbed. Their behavior in the visible region depends upon the value of the band gap. The appearances of such materials will range from the transparency of diamond or rocksalt (band zaps .. . 5 eV) to the metnllir 111s;erof p)~.itesor germnnrum (hand gaps less than I PVIV I R the adammtine luscer of the numerous . rthosr~hides . and sulfides which have intermediate values. For example, zinc sulfide has a hand gap of 3.58 eV (347 urn) which requires light in the nltraviolet region in order to promote electrons to the exrited state. Zinc sulfide cannot ahsorh visible light hecause visihle light is not energetic enough; therefore, zinc sulfide absorbs only ultraviolet light and reflects visible light. For the same reason, zinc oxide with a band gap of 3.2 e? is also white. On the other hand, cadmium sulfide, with a hand gap of 2.4 eV (517 nm) will ahsorh all electromagnetic radiation with an energy of 2.4 eV or greater. This corresponds to the entire blue reeion of the sunctrum. so the color reflected will be yellow, the cwnl~lemenrof hlue, and as we know, cadmium ~ r l f i d eis rndewl vrllow. On the other hand. cadmium selenide, with a hand g a i of around 1.8 eV, ahsorb; all visible light exrept the very low energy red region and thus appears red. Cadmium telluride, with a hand gap of 1.5eV, absorbs all visible lieht and thus amears hlack. These observations are .. s ~ m m a r ~ z eind Figure 2. Srminmrlurtor theories, especially thlnr involving rrysrsl defects, are c a p i ~ l & ~t,xplaining nf the

- ~.~.~ ~

~~~

-

P

Band Transitions

There is an additional set of compounds whose colors we have not yet explained, and these are materials like t,he cadmium compounds of the sulfur family. Since cadmium has a dlo electronic configuration, crystal field splitting rannot be the origin of cadmium sulfide's yellow color, hut the color can he explained on t,he basis of semiconductor properties. We are familiar wit,h the fact that any modification of the spacings between elect,ronic energy levds in atoms or mole482 1 Journal of Chemical Education

Valence Bond ZnO

White

CdS Yellow

CdSe

CdTe

Red

Black

Figure 2. Behavior of several semiconductingmaterials to visible radiation.

Although an electronic transition from one energy level to another is the process chiefly responsible for color in atoms, molecules, andcrystals, the fact that so many different hues and shades of color are observed in materials of the same chemical composition would lead one t o postulate that the enerev level differences are easily modified. Indeed, since an isolated atom can be energetically modified by greater proximity to ligands or t o like atoms, i t would he reasonable to assume that pure chemical compounds can he similarly modified by their surroundings. Listed below are only a few of the factors that can serve to modify the color of a given material, and therefore, factors to be reckoned with in the practical order.

chemical compounds as the examples cited below will illustrate. Turki and coworkers (37) and previous workers report that when the yellow form of mercury(I1) oxide is heated above room temperature, the compound turns red. Shelton (38) has observed that white copper(1) chloride becomes deep blue a t 178"C, and deepens to blue-black a t higher temperatures. then becomes " ereen-black and finally . -vields a deep green melt. On cooling, the sequence of colors is reversed. Shiriashi and colleagues a t the Matsushita Electric Company found that when they grew cadmium sulfide crystals a t different temperatures, different colors resulted. These color variations with temperature may have several different ex~lanations.One possibility is the existence of different crystalline forms at different tekperatures. Another is the possible presence of impurity centers which become more readily excited a t higher temperatures and thus more readily absorb lower energy electromagnetic radiation.

Crystal Form

Solvent or Disoersina Medium

Wegman (33) has observed that certain dyes, in spite of identical chemical composition, differ markedly in physical properties such as color and melting point if they exist in different crystalline modifications. A celebrated example is copper phthalocyanine which exists in two crystalline forms called a and (3. For several years, only the a-form was able to be prepared in suitable pigmentary form (34). Another notable e x a m ~ l eamone the inoreanic niements " . is lead chromate which can exist in the lemon-yellow rhomhic form, the reddish-yellow monoclinic form, and the scarlet tetragonal form. Only the monoclinic form is stable a t room temperature. hut the ~ r i m a r vorecioitation form is the vellow the rhombic fork. The ihombic form slowly converts monoclinic form, but it can be stabilized if coprecipitated with lead sulfate. Unless the rhomhic form is adequately stabilized, slow transformation to the monoclinic form will give rise to an orange-red nuance of tone and this can take place even in the dry pigment.

The fact that various coloring materials exhibit different colors in different solvents or dispersing media is well-known. This phenomenon is due to one, some, or all of the following factors: refractive index differential, variation of refractive index with wavelength, charge transfer interaction with the solvent. irreversible or reversible chemical reaction with the solventor dispersing medium, interaction hetween dye molecules themselves through association or polymerization on a suitable aulrstrnte, and selective scattering. For example. he phenommon of a yellow pigment yielding n green color upon gddition of ~ small~amoun& of carbon black . ~ ~ -~ ~ ~ - ~ - - ~ is a well-known result of some of these multiple interactions. If a thin acrylic coating containing a yellow pigment is applied over a hlack and white harkground, it will ahsorhstrongly in the blue (400 to 480 nm) and therefore will hnvp a higher index of refraction in thegreen (50010600 nm) than in the red (600 to700 nmJ reeion. The reason for this is thnt the refractive index maxim;m for chromatic materials usually occurs a t a slightly hipher waveleneth than the absor~tionmaximum. When c k t e d over white, the more efficient ;catwring of the 50n 6M nm lieht is nor, noticeable since the white background eventuallyreturns all unabsorbed light to the surf&e. However, the hlack background preferentially absorbs the less efficiently scattered red light, thus yielding only the efficiently scattered green light to the eye (36).

existence of color centers in normally colorless materials as well (32). Miscellaneous Color Production and Modification

Lattice Defects

Winter (35) has attributed the coloring of smoky quartz to the ioint effects of the existence of aluminum atoms. associated: with sodium and lithium atoms, in the lattice; and of neeative electrons and ~ o s i t i v eholes t r a o ~ e din interstitial He found that treatment by heat, X-rays or neutrons decolorized the auartz and that the change - in color was accompanied by a slight bluish luminescence. Pa~ticleSize and Shape

Light scattering by particles can be related to particle size. Theoretical considerations indicate that the particle size vieldine maximum scatter is of the order of the waveleneth bf lightkelf. Particles very much smaller than the wavelen2h of the incident light follow Rayleigh's law, which states that the intensity of the scattered light is very low and varies inversely with the fourth power of the wavelength and the sixth power of the particle diameter. Such particles are very efficient where transnarent colorations are desired. The cdor characteristics for; given pigment can thus be controlled by regulating the average particle size (36). Refractive Index

If the refractive index of a pigment is similar to that of the medium in which it is dispersed, then i t has a transparent quality which can he put to good use. For example, hlue copper phthalocyanine has a refractive index similar to the medium in which it is normally used and therefore gives transparent prints and lacquers. Use is made of this property when printing in four colon: yellow and red can be printed first, and the blue-black can be printed over it without dulling or masking the overall effect (34). Temperature

Changes in temperature seem to affect the color of pure

-

~~

~

Conclusion We have seen in this paper how important the chemical structure of a molecule is in determining its color or lack of it. We have seen that color arises from a t least five different types of chemical association capable of allowing electronic transitions in the visihle regionof the spectrum. Furthermore the enemetics of these rranaitions mn he mndified hy the physical composition and surroundings of the colored material. The colorants available today differ widely in their physical and chemical properties, hut they have one important feature in common: they all selectively absorb and scatter light in the to which the human eve res~onds.So we must s ~ e c t r areeion l never forget the fundamentally subjective nature of color. The ultimate reference with respect to all color perception is the human eye-brain system. The eye is an extraordinarily sensitive detector capable of distinguishing between five and eight million different colors (39), and the brain is a computer which makes the logical decisions. They function by dividing the overall spectral range into three distinct ranges, namely, red, green and hlue. Within each range, gradual color changes are perceived. Modification of colorants can modify this response, and we now know enough about the chemistry of colorants to explain why chlorophy~isgreen, rubies are red and chromate ion is vellow. ~xderiencehas shown that the prediction of the coloring properties of yet unsynthesized compounds is a very risky business which still remains in the realm of art rather than of Volume 55, Number 8, August 1978 1 483

science. So the world of color remains a mysterious and fascinating place, and the mystery only serves to deepen as we come to know more and more about it. Acknowledgment

I wish to thank Dr. Fred W. Billmeyer, Jr., who sograciously offered me the use of the facilities of the Rensselaer Color Measurement Laboratory during the preparation of this uauer. and the Facultv Fund of the Colleee of New Rochelle ?o; its financial supp&t.

Literature Clted (1) GerriLncn. F.. "Thmryand Practiceof Color." Van Noatmd-Reinhold Co.. New Vork. 1974.p~.9-18. (2) Committee on Colorimetryof the American Optical Society, "The Seienee of Color." American Optical Society. Washington. DC.. 1963.pp. 1644. (3) Hallas. G.. Sch. Sci. RPU.,56.507 (19751. 141 Rjermrn, N . , Z o n n g . Chem.. 63.140(1909). IS1 Billmoyer. F. W., .lr. and Saltlman, M., "Principles of Color Terhnolgy." Wiley-Inlersriancc. New York. 1966, pp. 2-12, (61 Srhultze. M..Alrh. Mtkl A n o t . 2.171(1&6). 17) Luria.S. M.. Physics Todw. 19.34 (March, 19661. 18) Wald,G.,Scimce, 145, IW7 (September. 1964). (9) MaeNichol, E. F., Jr., Scirnfihr Amsricon. 221,46 (December, 1964). (10) Hering, E.,"Zur Lehre vom Liehtsinne." Carl GemldZSohn, Wien, 1878. (111 Hawich, L. M.."TwDerade. nfopwnent Procps*..'in "Color 77, P r m d i w o f the Third Congress of the lnternationsl Color Asmriation: (Editor: Billmeyer. Jr.. F. W.) Adam Hilzer.Ltd.,London. 1978,pp.33-61. (12) C0ates.E.. J. Soc. Dyers Colour., 83.(1J.95(19671. (13) Wilf.0. N.,ZPhysik.Chsm., 9.522 (18761. (14) Jsffe, E, R a n d Mamhsll, W1,Chrmratry. 37.7 (Deember.1964). I151 Sandorfy,C.."ElectronieSpcctrasndQuantumChemiatry:Prentirr~Hall.Ine.,En-

484 1 Journal of Chemical Education

d d Cliffs, New Jereey. 1964, pp. 341-351. Jones, F., "The Colourand Constitution of Organic Moleeulea."in "Pigments: An IntmductlontoTheirPhysiealChemistry." (EdilacPatfemn:D.),AmericanElsevier Co.. New York, 1967, pp. 2649; Griffiths. J., "Color and Constitution of Organic Mol~ulea."Academie P-a, New Vork. 1976, pp. 1853. Mullikcn.R. S.. J Amer Chem. Soe, 74,811 (1952). Mulliken. R.S.. J. Phw. Chem., 56,801 (1952). McGl9mn.S.P.. Chem Re".. 58.1113(1954). Orgel. I.. E., Querlorly Reuicw (London), 8,422 (1954). Prribram. K..Z Phyaik.. 20.196 11928). Murrall, J . N.. Qumrlarly Reniema (London), 15.191 (1961). Rabinoaiteh, E.,RPu.Mod. Phya., 14,127 (1942). laver. A. R. P.. InorgonicEkclrnnicSpecfroaeopy, AmericanElaevier Co.. Nn* York, 196% ~~~-~

J$lpe"se", C. K.,"Absotpti~" 8-8 and Chcmieal B o d i l y in Camp1eXes.- P~'&M~o" Press. Inc.. New York, 1962. Robin. M. B. and Day. P.,.'Mixd Valenee Chemiltly." i n "Ad"."ees in lno,ganie Chemistryand Radiochemistry." Vol. 10, (Editon: Emeleus, H. J.. and Shsrpe, A. G.), AcadamiePress, New York, 1967,pp. 247422. Loeffler, B. M.,and Burns, R. G.. American Scientist. 64,638 (1916). Feitknecht, W.. "The Thmry of the Colour of Inorganic Substances." in "Pigment.: l (Editor Paitemon, D.). American An lntmduction to Their P h ~ i e a Chemiatry," Elrevier Co.. New York. 1967. pp. 1-25. Riehnrd8on.S. M..Am. Mineral, 60.73 (1975). Wornl, D. L..and Naaau. K., Am. Mineral.. 53.777 (1968). D., J.CHEM.EDUC.. Cotton, F. A,, J. CHEM. EDUC.,4I.466(1964);Liehr,Andr~~ 39.135 (1962); Gray, Hsmw R.,J. CHEM. EDUC.,41.2 11960:Companion, Audrey L., and Trevor, Dennis J.. J. CHEM. EDUC., 52,710 (19751. Suchet. J. P.. "Chemical Phyainof Semirondudors."D. Van Natrandand Co.,New IIe.6. YO& ~,~~~~~ Wsgman. .I.. J. Soe. Dyers Colour.. 76.282 (19MI). Smith, F. M.. J. Soc. Dyers Calour. 78.222 (1962). Winter,A..SilieaLsaInd.. 27.93 (1961). Papillo. P. J.. Modern Plorfiea, I39 (December. 1961). Turki, A. R., sf al., DokL Akod Nnub SSSR. 142,laJS 11962). She1ton.R. A. J . J . lnorg. Nucl Chem., 17.191 (19611. Huey. S., Color Eneine=ring. 9.20 (February, 1971). ~~