Chem. Rev. 1994f 94, 2319-2358
`
`2319
`
`Solvatochromic Dyes as Solvent Polarity Indicators
`
`Christian Reichardt
`
`Department of Chemistry and Scientific Centre for Material Sciences, Philipps University, Hans-Meerwein-Strasse, D-35032 Marbury, Germany
`
`Received March 23, 1994 (Revised Manuscript Received August 30, 1994)
`
`Contents
`2319
`I. Scope
`IL
`2319
`Introduction
`2321
`Ill. Linear Free-Energy Relationships
`2322
`IV. Solvatochromlsm
`A. Solvent Effects on UVNis/Near-lR Absorption 2322
`Spectra
`B. Solvatochromic Compounds
`V. Empirical Parameters of Solvent Polarity from
`UVNis/Near-lR Spectroscopic Measurements
`A. Single Parameter Approaches
`B. The Er(30) and ft; Scale of Solvent Polarity
`C. Multiparameter Approaches
`VL Interrelation between Empirical Solvent Polarity
`Parameters
`Vil. Summary and Conclusions
`vm. Acknowledgments
`IX. References
`
`2323
`2323
`
`2323
`2334
`2346
`2349
`
`2352
`2353
`2353
`
`I. Scope
`This review compiles positively and negatively
`solvatochromic compounds which have been used to
`establish empirical scales of solvent polarity by
`means of UV/vis/near-IR spectroscopic measurements
`in solution-with particular emphasis on the ET(30)
`scale derived from negatively solvatochromic pyri·
`dinium N-phenolate betaine dyes.
`This requires a short discussion of the concept of
`solvent polarity and how empirical parameters of
`solvent polarity can be derived and understood in the
`framework of linear free-energy relationships. The
`preconditions for the occurrence of solvatochromism,
`and further requirements of solvatochromic com(cid:173)
`pounds for them to be useful as solvent polarity
`indicators will be discussed. In addition to spectro(cid:173)
`scopically based single parameters of solvent polarity,
`multiparameter treatments of solvent effects by
`means of solvatochromic parameters will also be
`mentioned.
`The mutual interrelation between some of the more
`important tN/vis/near-IR spectroscopically derived
`solvent scales, and their correlations with solvato(cid:173)
`chromic multiparameter equations will be exemplar(cid:173)
`ily given.
`
`Introduction
`II.
`Rates and equilibrium positions of chemical reac(cid:173)
`tions, as well as the position and intensity of absorp(cid:173)
`tion bands in UV/vis/near-IR, IR, NMR, and ESR
`spectroscopy, are solvent-dependent.1- 15 Nowadays,
`this is generally known to every chemist, and the
`
`0009-2665/94/0794·2319$14.00/0
`
`Christian Reichardt was bom in 1934 in Ebersbach, Saxony, Germany.
`After a one-year stay 1953-1954 at the "Fachschule rur Energia" in Zittau,
`GDR, as teaching assistant, he studied chemistry at the ~earl Schorlem·
`mer" Technical University for Chemistry in Leuna-Merseburg, GDR,
`and-after moving illegally to West Germany in 1955-at the Philipps
`University in Marburg, FRG, where he obtained his Ph.~. in 1~?2 ~nd~r
`the tutelage of Professor K Dimroth, and completed his Hab1htat1on m
`1967. Since 1971 he has been Professor of Organic Chemistry at
`Marburg.
`In 1988 he was a visiting professor at the University of
`Barcelona, Spain. He has authored and co-authored more than 135
`papers and patents, and a book entitled Solvents and Sotven~ Effects in
`Organic Chemistry, which has been tran~!ated into ~rench, ~hmese, .and
`Russian. His research interests are m synthetic organic chemistry
`(chemistry of aliphatic dialdehydes, synthesis of polymethine dyes) and
`in physical organic chemistry (so!vatochromisrn of organic dyes, S?lvent
`effects in organic chemistry, empirical parameters of solvent polanty),
`
`careful selection of an appropriate solvent for a
`reaction or absorption under study is part of its
`craftsmen's skill. The influence of solvents on the
`rates of chemical reactions was first noted by Ber(cid:173)
`thelot and Pean de Saint-Gilles in 1862 in connection
`with their studies on esterification of acetic acid with
`ethanol ("The esterification is disturbed and deceler(cid:173)
`ated on addition of neutral solvents not belonging to
`the reaction>')16 and followed by the pioneering work
`of Menshutkin in 1890 on the alkylation of tertiary
`amines with haloalkanes. 17•18 Menshutkin's state(cid:173)
`ment that "a chemical reaction cannot be separated
`from the medium in which it is performed" still
`remains valid-and has recently been more casually
`expressed as "In searching to understand the rate of
`a reaction in solution, the baby must not be separated
`from its bath water". 19
`The influence of solvents on the position of chemi(cid:173)
`cal equilibria was discovered in 1896 by Claisen, 20
`Knorr, 21 and Wislicenus22 independently of each
`other, simultaneously with the discovery ofketo-enol
`tautomerism in 1/ ··
`wrote, " ... It depen
`FINCHIMICA EXHIBIT 2027
`the temperature, ~
`on the nature of tl AD AMA MAKHTESHIM v. FINCHJMICA
`CASE IPR2016-00577
`
`© 1994 American Chemical
`
`

`
`Chem. Rev. 1994, 94, 2319-2358
`
`2319
`
`Solvatochromic Dyes as Solvent Polarity Indicators
`
`Christian Reichardt
`
`Deparlment of Chemistry and Scientific Centre for Material Sciences, Philipps University, Hans-Meerwein-Strasse, D-35032 Marburg, Germany
`
`Received March 23, 1994 (Revised Manuscript Received August 30, 1994)
`
`Contents
`2319
`I. Scope
`2319
`II. Introduction
`2321
`Ill. Linear Free-Energy Relationships
`2322
`IV. Solvatochromism
`A. Solvent Effects on UVNis/Near-lR Absorption 2322
`Spectra
`B. Solvatochromic Compounds
`V. Empirical Parameters of Solvent Polarity from
`UVNis/Near-lR Spectroscopic Measurements
`A. Single Parameter Approaches
`B. The Er(30) and ~ Scale of Solvent Polarity
`C. Multiparameter Approaches
`VI. Interrelation between Empirical Solvent Polarity
`Parameters
`VII. Summary and Conclusions
`VIII. Acknowledgments
`IX. References
`
`2323
`2334
`2346
`2349
`
`2352
`2353
`2353
`
`2323
`2323
`
`Christian Reichardt was born in 1934 in Ebersbach, Saxony, Germany.
`After a one-year stay 1953--1954 at the "Fachschule fur Energia" in Zittau,
`GDR, as teaching assistant, he studied chemistry at the "Carl Schorlem(cid:173)
`mer" Technical University for Chemistry in Leuna-Merseburg, GDR,
`and-after moving illegally to West Germany in 1955- at the Philipps
`University in Marburg, FRG, where he obtained his Ph.D. in 1962 under
`the tutelage of Professor K. Dimroth, and completed his Habilitation in
`1967. Since 1971 he has been Professor of Organic Chemistry at
`Marburg.
`In 1988 he was a visiting professor at the University of
`Barcelona, Spain. He has authored and co-authored more than 135
`papers and patents, and a book entitled Solvents and Solvent Effects in
`Organic Chemistry, which has been translated into French, Chinese, and
`Russian. His research interests are in synthetic organic chemistry
`(chemistry of aliphatic dialdehydes, synthesis of polymethine dyes) and
`in physical organic chemistry (solvatochromism of organic dyes, solvent
`effects in organic chemistry, empirical parameters of solvent polarity).
`
`careful selection of an appropriate solvent for a
`reaction or absorption under study is part of its
`craftsmen's skill. The influence of solvents on the
`rates of chemical reactions was first noted by Ber(cid:173)
`thelot and Pean de Saint-Gilles in 1862 in connection
`with their studies on esterification of acetic acid with
`ethanol ("The esterification is disturbed and deceler(cid:173)
`ated on addition of neutral solvents not belonging to
`the reaction")16 and followed by the pioneering work
`of Menshutkin in 1890 on the alkylation of tertiary
`· amines with haloalkanes. 17•18 Menshutkin's state(cid:173)
`ment that "a chemical reaction cannot be separated
`from the medium in which it is performed" still
`remains valid-and has recently been more casually
`expressed as "In searching to understand the rate of
`a reaction in solution, the baby must not be separated
`from its bath water". 19
`The influence of solvents on the position of chemi(cid:173)
`cal equilibria was discovered in 1896 by Claisen,20
`Knorr,21 and Wislicenus22 independently of each
`other, simultaneously with the discovery of keto- enol
`tautomerism in 1,3-dicarbonyl compounds. Claisen
`wrote, " .. .It depends on the nature of substituents,
`the temperature, and for dissolved compounds, also
`on the nature of the solvent, which of the two forms
`
`I. Scope
`This review compiles positively and negatively
`solvatochromic compounds which have been used to
`establish empirical scales of solvent polarity by
`means of UV/vis/near-IR spectroscopic measurements
`in solution-with particular emphasis on the ET(30)
`scale derived from negatively solvatochromic pyri(cid:173)
`dinium N-phenolate betaine dyes.
`This requires a short discussion of the concept of
`solvent polarity and how empirical parameters of
`solvent polarity can be derived and understood in the
`framework of linear free-energy relationships. The
`preconditions for the occurrence of solvatochromism,
`and further requirements of solvatochromic com(cid:173)
`pounds for them to be useful as solvent polarity
`indicators will be discussed. In addition to spectro(cid:173)
`scopically based single parameters of solvent polarity,
`multiparameter treatments of solvent effects by
`means of solvatochromic parameters will also be
`mentioned.
`The mutual interrelation between some of the more
`important UV/vis/near-IR spectroscopically derived
`solvent scales, and their correlations with solvato(cid:173)
`chromic multiparameter equations will be exemplar(cid:173)
`ily given.
`
`II. Introduction
`Rates and equilibrium positions of chemical reac(cid:173)
`tions, as well as the position and intensity of absorp(cid:173)
`tion bands in UV/vis/near-IR, IR, NMR, and ESR
`spectroscopy, are solvent-dependent.1- 15 Nowadays,
`this is generally known to every chemist, and the
`
`0009-2665/94/0794-2319$14.00/0
`
`© 1994 American Chemical Society
`
`

`
`2320 Chemical Reviews, 1994, Vol. 94, No. 8
`
`(i.e. keto and enol form) will be the more stable."22
`These results were first reviewed by Stobbe in 1903,23
`who divided the solvents used into two groups ac(cid:173)
`cording to their ability to isomerize tautomeric
`compounds. To some extent, his classification reflects
`the modern division of solvents into hydrogen-bond
`donor (HBD, protic) solvents and non-hydrogen-bond
`donor (non-HBD, aprotic) solvents.
`In contrast to these more historical investigations,
`a few recent examples from different areas shall
`demonstrate the powerful influence of solvents on
`chemical reactions and spectral absorptions:
`(a) The equilibrium constant of the 1:1 complex
`formed between a macrobicyclic cyclophane receptor
`and pyrene varies by a factor of ca. 106 upon changing
`the solvent from carbon disulfide to water, which
`corresponds to a solvent-induced difference in the
`Gibbs binding energy of Ll!lG 0 = 8.1 kcaVmol.24,25
`(b) Comparison of the unimolecular heterolysis rate
`constants of 2-chloro-2-methylpropane obtained in
`benzene and in water reveals a rate acceleration of
`ca. 1011 with increasing solvent polarity.26,27 The
`first-order rate constant of the decarboxylation of
`6-nitrobenzisoxazole-3-carboxylate varies by up to 8
`orders of magnitude on going from reaction in hexa(cid:173)
`methylphosphoric triamide to reaction in water. 28·29
`(c) The intramolecular charge-transfer UV/vis/near(cid:173)
`IR absorption band of the solvatochromic 2,6-diphen(cid:173)
`yl(2,4,6-triphenyl-1-pyridinio)phenolate betaine dye
`36 is shifted from Amax = 810 nm to Amax = 453 nm
`(Ll?i. = 357 nm, Llv = 9730 cm-1) on going from
`diphenyl ether to water as solvent.1 This corresponds
`to a solvent-induced change in excitation energy of
`ca. 28 kcaVmol.
`(d) In the fluorescence spectrum of 1-phenyl-4-[(4-
`cyano-1-naphthyl)methylene]piperidine (71), the emis(cid:173)
`sion maximum is shifted from Amax = 407 nm to Amax
`= 694 nm (Ll?i. = 287 nm, Llv = 10200 cm-1) by
`changing the solvent from n-hexane to acetonitrile.30
`(e) The solvent-induced IR frequency shift for the
`C=O stretching vibration of tetramethylurea is Llv
`= 71 cm-1 in going from n-hexane (v = 1656 cm-1) to
`water (v = 1585 cm-1) as solvent.31 Corresponding
`solvent effects on the IR spectra of ethyl acetate [v(cid:173)
`(C=O)] and acetonitrile [v(C=N)] have been very
`recently obtained.32•33
`(f) The solvent-induced difference in the 31P NMR
`chemical shift oftriethylphosphane oxide, measured
`inn-hexane and in water, is !lo~ 23 ppm.34,35 Even
`the NMR chemical shift of the non polar solute 129Xe
`can vary by up to !lo = 200 ppm depending on the
`solvent used. 36·37
`(g) The nitrogen and hydrogen hyperfine splitting
`constants, as well as the spin densities on the
`nitrogen and carbon atoms, taken from the ESR
`spectrum of the 2-[ 4-(dimethylamino)phenyl]indan-
`1,3-dionyl radical, are highly solvent-dependent.38
`Similar solvent effects on ESR spectra have also been
`found recently with paramagnetic organometallic
`complexes such as Co(COhL2 (L = chelating phos(cid:173)
`phane).39
`Responsible for all these medium effects is the
`differential solvation of (i) reactants and products ((cid:173)
`position of chemical equilibria); (ii) reactants and
`activated complexes (- rates of chemical reactions);
`
`Reichardt
`
`or (iii) molecules in the corresponding ground and
`excited states (- physical absorption of electromag(cid:173)
`netic radiation). The extent of this differential sol(cid:173)
`vation depends on the intermolecular forces between
`solute and surrounding solvent molecules. Intermo(cid:173)
`lecular forces include nonspecific forces such as
`purely electrostatic forces arising from the Coulomb
`forces between charged ions and dipolar molecules
`[i.e. ion/ion, ion/dipole, dipole/dipole] and polarization
`forces that arise from dipole moments induced in
`molecules by nearby ions or dipolar molecules [i.e.
`ion/nonpolar molecule, dipole/nondipolar molecule,
`two nonpolar molecules (dispersion energy)], as well
`as specific forces such as hydrogen-bonding between
`HBD and HBA ions or molecules, and electron-pair
`donor (EPD)/electron-pair acceptor (EPA) forces. 40-42
`Obviously, intermolecular solute/solvent interactions
`are of highly complicated nature and difficult to
`determine quantitatively.
`Chemists have tried to understand solvent effects
`on chemical reactions in terms of the so-called solvent
`polarity, which is not easy to define and to express
`quantitatively. What does solvent polarity mean?
`The simplicity of idealized electrostatic models for the
`description of solvation of ions and di polar molecules,
`considering solvents as nonstructured continuum,
`has led to the use of physical constants, such as static
`dielectric constant (E,), permanent dipole moment(µ),
`refractive index (n), or functions thereof, as macro(cid:173)
`scopic solvent parameters for the evaluation of me(cid:173)
`dium effects. However, solute/solvent interactions
`take place on a molecular microscopic level within a
`structured discontinuum consisting of individual
`solvent molecules, capable of mutual solvent/solvent
`interactions. For this reason, and because of neglect(cid:173)
`ing specific solute/solvent interactions, the electro(cid:173)
`static approach to medium effects often failed in
`correlating observed solvent effects with physical
`solvent parameters.1 In reality, satisfactory quan(cid:173)
`titative descriptions of medium effects have to take
`into account all nonspecific and specific solute/
`solvent, solvent/solvent and, at higher concentrations,
`even solute/solute interactions. Therefore, from a
`more pragmatic point of view, it seems to be more
`favorable to define "solvent polarity" simply as the
`overall solvation capability (or solvation power) of
`solvents, which in turn depends on the action of all
`possible, nonspecific and specific, intermolecular
`interactions between solute ions or molecules and
`solvent molecules, excluding, however, those interac(cid:173)
`tions leading to definite chemical alterations of the
`ions or molecules of the solute (such as protonation,
`oxidation, reduction, chemical complex formation,
`etc.). This definition of solvent polarity was given
`in 1965,1·43 and it seems to be becoming more and
`more accepted by the scientific community.44- 46
`Apparently, solvent polarity thus defined cannot
`be described quantitatively by single physical solvent
`parameters such as dielectric constants, dipole mo(cid:173)
`ments, etc. The lack of comprehensive theoretical
`expressions for the calculation or prediction of solvent
`effects on chemical reactivity, and the inadequacy of
`defining solvent polarity in terms of simple physical
`solvent characteristics, have led to the introduction
`of so-called empirical parameters of solvent polar-
`
`

`
`Solvatochromic Dyes
`
`Chemical Reviews, 1994, Vol. 94, No. 8 2321
`
`ity. 1,4s.41 On the basis of the assumption that par(cid:173)
`ticular, carefully selected, well-understood and strongly
`solvent-dependent chemical reactions or spectral
`absorptions may serve as suitable model processes
`for recording medium effects; various empirical sol(cid:173)
`vent polarity scales have been developed this way.1,47
`The desmotropic constant, L, introduced by Meyer
`in 1914 as a measure of the enolization power of
`solvents for 1,3-dicarbonyl compounds, can be con(cid:173)
`sidered as the first empirically determined solvent
`parameter, using the keto-enol tautomerization of
`ethyl acetoacetate as the solvent-dependent reference
`process.48 However, the first real empirical param(cid:173)
`eter of "solvent ionizing power" was the Y scale
`introduced by Winstein et al. in 1948, derived from
`the SNl heterolysis of2-chloro-2-methylpropane.47g,49
`The first suggestion that solvatochromic dyes could
`serve as visual indicators of solvent polarity was
`made by Brooker et al. (from the Eastman Kodak
`Company in Rochester, NY) in 1951,50 but Kosower
`was the first to set up a real spectroscopic solvent
`polarity scale in 1958. This was called the Z scale
`and used the intermolecular charge-transfer (CT)
`absorption of 1-ethyl-4-(methoxycarbonyl)pyridinium
`iodide as the solvent-sensitive reference process.3,44,51
`Since then, various additional UV/vis/near-IR-based
`solvent polarity scales have been developed, using
`negatively or positively solvatochromic dyes of dif(cid:173)
`ferent chemical structure, and, depending on their
`structure, capable of registering all or only selected
`types of intermolecular dye(solute)/solvent interac(cid:173)
`tions.1·47 The main aim of this paper is to collect
`these solvatochromically derived solvent polarity
`scales and to compare them with respect to their
`usefulness.
`In applying such single-parameter solvent scales,
`it is tacitly assumed that the combination of solute/
`solvent interactions between the reference solute(s)
`and the solvent is almost the same as with the
`particular substrate under consideration. Obviously,
`this is an oversimplification which causes serious
`limitations of the single-parameter approach to me(cid:173)
`dium effects. Therefore, more recently, multiparam(cid:173)
`eter correlation equations have been developed,
`which consist of up to four single empirical param(cid:173)
`eters, each of them measuring a certain aspect of the
`overall solvation capability of a given solvent (e.g.
`solvent polarizability, dipolarity, Lewis acidity, and
`Lewis basicity).1,47a,b.f,52·53 If the one-parameter ap(cid:173)
`proach for correlating solvent effects fails, then
`multiparameter correlations come into play.
`This method of proceeding, i.e. the use of reference
`or standard compounds in order to establish empiri(cid:173)
`cal solvent polarity parameters, is quite common in
`chemistry and takes usually the form of a linear free(cid:173)
`energy (LFE) relationship.54-57
`
`Ill. Linear Free-Energy Relationships
`LFE relationships involve empirical relationships
`between rates or equilibria of chemical reactions,
`which show some similarity within a so-called reac(cid:173)
`tion series. Considering a chemical reaction between
`a substrate S and a reagent R in a medium M, which
`leads, via an activated complex, to the product(s) P,
`according to
`
`there are three possibilities of introducing small
`changes in order to establish a reaction series:1,54-57
`(a) First, one can change the substrate by introduc(cid:173)
`ing different substituents. This leads, particularly
`in case of meta- and para-substituted benzene de(cid:173)
`rivatives, to the well-known Hammett equation.58 A
`recent, typical example of this kind of LFE relation(cid:173)
`ship is the substituent-dependent alkylation of py(cid:173)
`ridinium N-phenolate betaine dyes.59,60
`(b) Second, one can change the reagent (equal to
`catalyst). This gives, e.g. in case of acid- or base(cid:173)
`catalyzed reactions, the famous Brt11nsted-Pedersen
`equation, which establishes a LFE relationship be(cid:173)
`tween the strength of acids or bases and their
`effectiveness as catalysts. 61 This catalysis equation,
`first introduced in 1924, was the first LFE relation(cid:173)
`ship.
`(c) Third, in order to obtain a reaction series, one
`can change the surrounding medium, while leaving
`all other reaction partners unchanged. In the case
`of sufficiently solvent-dependent chemical reactions,
`this leads to kinetically derived empirical parameters
`of solvent polarity, such as the Yvalues ofWinstein.49
`A simple modification to the previous equation, by
`replacing the reagent R with photons hv and the
`product P with the substrate Sin the spectroscopi(cid:173)
`cally excited state, leads to
`
`(S)M + hv -
`ground
`state
`
`(S)~
`excited
`state
`
`This replacement now corresponds to an extension
`of the LFE principle, as applied to reaction series, to
`so-called absorption series, which are available in all
`areas of absorption spectroscopy (UV /vis/near-IR, IR,
`ESR, NMR). In order to establish an absorption
`series, there are two possibilities to alter the param(cid:173)
`eters of this equation:
`(a) First, one can again change the substrate by
`introducing different substituents. This leads to
`spectroscopically derived Hammett equations con(cid:173)
`necting substituent-induced wavenumber shifts of
`suitably substituted substrates with Hammett sub(cid:173)
`stituent constants. Examples of Hammett relation(cid:173)
`ships for UV/vis spectroscopic data can be found in
`the literature. 62·63
`(b) Second, only the medium in which the substrate
`is to be dissolved is changed. Provided the position
`of the spectral absorption band of the substrate is
`sufficiently solvent dependent, this procedure can be
`used to establish spectroscopically derived scales of
`solvent polarity, as described in this paper for UV/
`vis/near-IR absorptions. That is, solvent polarity
`scales as derived by means of absorption spectroscopy
`and their applications are, in principle, further
`examples of LFE relationships. 1,54- 57 However, the
`important question of whether LFE relationships are
`fundamental laws of chemistry, 64 or only locally valid,
`empirical rules,65 is still a matter of debate.
`
`

`
`2322 Chemical Reviews, 1994, Vol. 94, No. 8
`
`IV. Solvatochromism
`A. Solvent Effects on UVNis/Near-lR Absorption
`Spectra
`It has long been known that UV/vis/near-IR ab(cid:173)
`sorption spectra of chemical compounds may be
`influenced by the surrounding medium and that
`solvents can bring about a change in the position,
`intensity, and shape of absorption bands.66-68 Hantz(cid:173)
`schlater termed this phenomenon solvatochromism.69
`However, the now generally accepted meaning of the
`term solvatochromism differs from that introduced
`by Hantzsch. 1 One of the referees has recommended
`to replace solvatochromism by the term perichromism
`(from Greek peri = around) in order to stress that
`spectroscopic probe molecules cannot only measure
`the polarity of liquid environments, but also that of
`solids, glasses, and surfaces. The term solvato(cid:173)
`chromism is, however, so well established in the
`literature that it would be difficult to convince the
`scientific community to change this term to peri(cid:173)
`chromism, which is certainly a more general expres(cid:173)
`sion for the spectroscopic phenomena under consid(cid:173)
`eration.
`A hypsochromic (or blue) shift of the UV/vis/near(cid:173)
`IR absorption band, with increasing solvent polarity
`is usually called "negative solvatochromism". The
`corresponding bathochromic (or red) shift, with in(cid:173)
`creasing solvent polarity, is termed "positive solva(cid:173)
`tochromism". Obviously, solvatochromism is caused
`by differential solvation of the ground and first
`excited state of the light-absorbing molecule (or its
`chromophore): if, with increasing solvent polarity,
`the ground-state molecule is better stabilized by
`solvation than the molecule in the excited state,
`negative solvatochromism will result. Or vice versa,
`better stabilization of the molecule in the first excited
`state relative to that in the ground state, with
`increasing solvent polarity, will lead to positive
`solvatochromism. In this context, "first excited state"
`means the so-called Franck-Condon excited state
`with the solvation pattern present in the ground
`state.
`Since the time required for a molecule to get
`electronically excited (about 10-15 s) is much shorter
`than that required to execute vibrations or rotations
`(about 10-12 to 10-10 s), the nuclei of the absorbing
`entity (i.e. absorbing molecule + solvation shell) do
`not appreciably alter their positions during an elec(cid:173)
`tronic transition (Franck-Condon principle).70 There(cid:173)
`fore, the first excited state of a molecule in solution
`has the same solvation pattern as the corresponding
`ground state and is called Franck-Condon excited
`state, whereas the ground state corresponds to an
`equilibrium ground state. If the lifetime of the
`excited molecule is large enough, then reorientation
`of the solvent molecules, according to the new excited
`situation, takes place, and a relaxed excited state
`with a solvent shell in equilibrium with this state
`results. It is from this equilibrium excited state that
`fluorescence can occur. By analogy, there is a
`Franck-Condon ground state after emission with the
`solvation pattern of the equilibrium excited state,
`which persists briefly until the solvent molecules
`reorganize to the equilibrium ground state. The
`
`Reichardt
`
`differential solvation of these two states is respon(cid:173)
`sible for the solvent influence on emission or fluo(cid:173)
`rescence spectra. The solvent dependence of the
`position of emission bands in fluorescence spectra has
`been often included in the term solvatochromism. 1
`The solvent dependence of fluorescence spectra has
`been sometimes called solvatofiuorchromism 71 or
`fiuorosolvatochromism. 72 However, because of the
`close connection between spectral absorption and
`emission, there is no need for special terms for
`fluorescence-based solvatochromism.
`The solvatochromism observed depends on the
`chemical structure and physical properties of the
`chromophore and the solvent molecules, which, for
`their part, determine the strength of the intermo(cid:173)
`lecular solute/solvent interactions in the equilibrium
`ground state and the Franck-Condon excited state.
`This is not the place to discuss the relation between
`extent and sign of solvatochromism and the structure
`of solvatochromic dyes; the reader is referred to
`recent reviews.1·73-80 In general, dye molecules with
`a large change in their permanent dipole moment
`upon excitation exhibit a strong solvatochromism. If
`the solute dipole moment increases during the elec(cid:173)
`tronic transition {µg < µe), a positive solvatochromism
`normally results. In the case of a decrease of the
`solute dipole moment upon excitation {µg > µe), a
`negative solvatochromism is usually observed. Sol(cid:173)
`utes with this particular solvatochromic behavior can
`be commonly found among so-called meropolyme(cid:173)
`thine dyes (particularly among merocyanine dyes =
`vinylogous amides) and among compounds with
`inter- or intramolecular CT absorptions.1·73•77,78 In
`addition to the dipole moment change on excitation,
`the ability of a solute to donate or to accept hydrogen
`bonds to or from surrounding solvent molecules in
`its ground and Franck-Condon excited state deter(cid:173)
`mines further the extent and sign of its solvato(cid:173)
`chromism.81-86 Some merocyanine dyes (e.g. dye 48
`in Table 150·197) even show an inverted solvato(cid:173)
`chromism, i.e. their long wavelength solvatochromic
`absorption band exhibits first a bathochromic and
`then a hypsochromic band shift as the solvent polar(cid:173)
`ity increases. This is due to a solvent-induced change
`of the electronic ground-state structure from a less
`dipolar (in nonpolar solvents) to a more dipolar
`chromophore (in polar solvents) with increasing
`solvent polarity.197
`The search for quantitative relationships between
`the solvent influence on UV/vis/near-IR spectra and
`physical solvent parameters led Kundt, in 1878, to
`propose the rule, later named after him, that increas(cid:173)
`ing dispersion (i.e. increasing index of refraction) of
`the solvent results in bathochromic shifts of the
`solute absorption band. 66 Since then, numerous
`quantitative relationships between solute light ab(cid:173)
`sorption and physical solvent properties, based on
`different models for solute/solvent interactions (such
`as, for example, the Onsager reaction field approach)
`have been established. The discussion of these
`relationships is outside the scope of this review, and
`the reader is referred to a selection of publications
`dealing with various theoretical treatments of solvent
`effects on electronic spectra. 74,75,78,79,87-102
`
`

`
`Solvatochromic Dyes
`
`Chemical Reviews, 1994, Vol. 94, No. 8 2323
`
`The complexity of intermolecular solute/solvent
`interactions has led to correspondingly complex,
`theoretically derived relationships between solvent(cid:173)
`induced band shifts and physical parameters of solute
`and solvent, which, in general practice, have been
`rather seldom used by chemists in their efforts to
`quantify the term "solvent polarity". The main
`shortcomings of the theoretical treatments of solvent
`effects on electronic spectra are the unavoidable use
`of simplified model concepts, without due regard to
`the specific solute/solvent interactions such as hy(cid:173)
`drogen-bonding, EPD/EPA, and solvophobic interac(cid:173)
`tions. The lack of reliable theoretical calculations of
`solvent effects in the past, and the inadequacy of
`defining "solvent polarity" in terms of simple physical
`solvent constants, have stimulated attempts to in(cid:173)
`troduce empirical scales of solvent polarity, based on
`convenient, well-known, easily measurable, solvent(cid:173)
`sensitive reference processes within the framework
`of LFE relationships. 54- 57
`
`B. Solvatochromic Compounds
`Because of the simplicity of UV/vis/near-IR spec(cid:173)
`troscopic measurements, empirical parameters of
`solvent polarity have been preferably determined by
`means of solvatochromic compounds. It is assumed
`that a particular solvent-influenced UV/vis/near-IR
`absorption is a suitable, representative model for a
`large class of other solvent-dependent processes.
`Model processes used to establish spectroscopically
`empirical scales of solvent polarity have been re(cid:173)
`viewed.1,43,47.54 Solvatochromic compounds suitable
`as color indicators for solvent polarity measurements
`have also been reviewed. 1,103- 109 It should be noted
`that the absorption range of suitable solvatochromic
`reference compounds does not only include the tra(cid:173)
`ditional UV and vis region, but also the near-IR
`region. 10
`In Table 1 solvatochromic compounds, which have
`been used as UV/vis/near-IR spectroscopic indicators
`to establish empirical scales of solvent polarity, are
`compiled. They are roughly ordered according to
`their solvatochromic range, i.e. their sensitivity to a
`solvent change. Sometimes, these scales have been
`given a special name or symbol, after the type of the
`respective light absorption (:rr -
`:rr*, CT, etc.); this
`symbol is added in parentheses. Included in Table
`1 are also some of those solvatochromic compounds
`which have been proposed as solvent polarity indica(cid:173)
`tors, for which, however, a complete solvent scale has
`not been worked out. Naturally, there are many
`more known solvatochromic compounds than those
`given in Table 1. It is not easy to draw a distinct
`line between plain solvatochromic compounds and
`those which have been occasionally proposed as
`potential solvent polarity indicators. For a particular
`solvatochromic compound of Table 1, most of the
`relev