SEARCH

SEARCH BY CITATION

Keywords:

  • circular dichroism;
  • single crystals;
  • Nujol Mulls;
  • pellets;
  • diffuse reflectance;
  • suspensions;
  • films;
  • absorption flattening;
  • scattering

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION OF THE LITERATURE AND RESULTS
  5. CONCLUSIONS
  6. LITERATURE CITED
  7. APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

The interest of circular dichroism in the solid state is stimulated by several needs, such as the desire to get solvent free spectra, the insolubility of the sample or the intrinsic process in which the sample itself is prepared or manipulated. We approach the argument on the basis of the sampling technique, since each different case calls for specific care in getting proper results. Chirality 21:E28–E36, 2009. © 2009 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION OF THE LITERATURE AND RESULTS
  5. CONCLUSIONS
  6. LITERATURE CITED
  7. APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

Lately there has been a revived interest in the properties of chiral materials in the solid state. Scientists employing different techniques, like x-ray crystallography,1 dynamic light scattering,2 or from different areas like spectroscopy and separation science3 have dealt with optical active aggregates of chiral molecules in the solid state. Yet the main techniques are optical rotatory dispersion and circular dichroism (CD), and scientists listed above are aware of this and ask for help and collaboration from experts in the field: this is what has motivated us in writing this review. Besides Pescitelli and coworkers have recently reported that solid-state CD on natural products shows advantages with respect to solution phase CD, when comparing with TDDFT calculations, since the number of conformations is reduced to the solid-state structure.4, 5

A previous review on solid-state CD had been published 10 years ago in this journal by Reiko Kuroda.6 The same author gave further important contributions on this subject with other extended texts in different books edited from 2000 to 2008.7–9

Revisiting today the same argument gives us an opportunity to present a few further developments, but at the same time it allows us to extend the interest originally focused on inorganic samples. In addition today CD is no longer limited to the UV-vis range (ECD) since CD in the IR range (VCD) has become an established, routine technique: here too solid-state sampling may be important.

CD spectroscopy has been mainly applied to samples in solution phase, which means perfectly isotropic samples. On the contrary, solid state samples are often far from ideal in this respect. When using general purpose commercial CD spectrometers, potential artifacts may be present in the obtained spectra according to the type of sample and the selected sampling technique. We will therefore review here the topic of solid-state CD following the different sampling methods, while in the appendix we will discuss the most used theoretical treatment on the origin of anisotropy artifacts from conventional spectrometers.

DISCUSSION OF THE LITERATURE AND RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION OF THE LITERATURE AND RESULTS
  5. CONCLUSIONS
  6. LITERATURE CITED
  7. APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

Single Crystals

The ability to obtain reliable CD spectra on single crystals along directions other than the optical axis has been a difficult task for many years. The target was achieved by a dedicated spectrometer (J-800KCM) based on an extensively modified commercial unit,10, 11 which is suitable only for relatively large sized crystals.

The idea of a dedicated CD microscope was tried in the past,12 with questionable results, but the dream has been made possible by the innovative development of a CD imaging microscope,13–15 the idea behind this has also been patented.16

Both approaches, together with the sister HAUP technique17–19 (High Accuracy Universal Polarimetry) make use of a rather different hardware from the common CD spectrometers. The same is true for the promising thermal lens microscope which has not been yet applied on single crystals.20 On the other hand extensive modifications to a commercial spectrometer (JASCO J-820) has recently allowed CD applications on solid samples (60 × 60 μm size) with CCD imaging of the selected sampling area.21

Nujol Mulls

The Nujol mull technique is widely used in IR and less often in UV-vis absorption spectroscopies: a thin layer of Nujol (refined mineral oil) film sandwiched between suitable optical windows is transparent all over the UV-vis range down to well below 200 nm and in most of the IR range. Fluorolube (perfluorohydrocarbon) is occasionally used as Nujol alternative, but is far less transparent in the UV region.

With some care during sample preparation, finely ground chiral nonracemic powders easily give fair CD spectra in Nujol mulls: Figure 1 shows the CD spectra of Binaphtol enantiomers, quality mirror image spectra can be quickly obtained, since the finely dispersed sample may be considered of isotropic nature. The main drawback of the technique is reported to be scattering, a way to compensate had been proposed in the past22, 23 using the Shibata opal glass technique,24 which calls for placing a special diffuser just after the sample to homogenize the effect. A good alternative practice is to put the sample as close as possible to the photomultiplier tube surface (typically much larger than sample beam) to increase the solid angle collection and, in instruments with focused sampling geometry, to place the sample where the beam size is large enough to compensate patchy sample distribution. With both these approaches we have however no way to correct absorption flattening (AF) coming from inhomogeneous particles distribution in the mull. As a consequence the simplest approach is to run several spectra at various sample concentrations, since AF is strictly related to sample absorbance.25 In any case the real main “diagnostic” point is to collect the absorption spectra together with the CD data, directly or through the conversion of the high voltage applied to the photomultiplier tube. Any distortion in the absorption spectra will indicate either AF effects influencing also CD intensity or abnormal scattering typically limiting the signal-to-noise ratio of the CD data. In the absorption spectra the two effects go into different directions: a larger scattering will produce apparent higher absorbance, while a larger AF will depress the absorption intensity. These simple rules apply to UV-vis CD experiments, while in the IR range there is no way to increase the solid angle collecting geometry given the small surface of the detectors, but in this region scattering artifacts are much lower. Properly obtained Nujol mulls from randomly oriented particulate systems allow, in quite a few instances, direct comparison between solid state and solution spectra and several examples have been published in the UV-vis26–29 and IR field.30

thumbnail image

Figure 1. CD spectra (top) and UV spectra (bottom) of (S) and (R) enantiomers of 1,1′-binaphtols in Nujol in black and gray respectively. Notice that different absorption intensities correspond fairly to different CD intensities. Spectra taken with a JASCO J-815SE at the University of Brescia, with 1 nm SBW, 1 sec response and 50 nm/min scanning speed.

Download figure to PowerPoint

An unusual application of solid state CD in Nujol mulls has been reported under high pressure,31 these experiments are however very challenging due to the potential induced artifacts.

Pellets

There are cases in which the sample is dissolved or is chemically modified when it is put in contact with mineral oil. An alternative preparation method is based on mixing the sample with an inert matrix and for pressing UV-VIS/IR transparent pellets. KBr, CsI and KCl are the most frequently used “inert” materials, the latter being more UV transparent than the most used KBr. Pellet preparation requires a minimum of practice. Samples must be mixed with KBr or other transparent salts at a concentration of few percents and mixed so as to obtain a particle size ideally lower then the measuring wavelength. The mixture must then be pressed in an evacuable die (to remove humidity and to obtain transparent pellets with minimum scattering) equipped with separate anvils (in order not to rotate the powder when pressure is applied, since this may “orient” the sample). The obtained pellet includes two components diffusing the incoming light: the sample itself and the salt, so the Shibata opal glass approach suggested for mulls is in this case performed by the grinded salt itself.

As in the case of Nujol mulls, pellets should be placed in the UV-vis CD spectrometers as close as possible to the photomultiplier tube surface; in addition some users rotate pellets around the light propagation axis at discrete angles to compensate for non constant pellet thickness/surface parallelism and take the average of the different runs. Some other users average the reading continuously by spinning the pellet during the spectra acquisition.

In the pellet technique too the sample is in isotropic conditions and artifacts from spurious linear polarization components are largely excluded, but AF effects may well be present so, as outlined above, it's essential to extract also the regular absorption data to verify the consistency of the measurement.

Pellet techniques have been used in many laboratories, for a variety of samples including biomolecules,32–43 both in the UV-vis (ECD) and in the IR (VCD).

IR plates

Commercially available alkali halide sample support substrates (International Crystal Laboratories- Garfield, NJ) have been proposed in NaCl, KBr or KCl variants for fast qualitative IR analysis for liquids, mulls, film casting from solutions. These can be used in ECD and VCD too: a few drops of a regular VCD solution of (−)-Camphor in CCl4 were dispersed on the NaCl surface and solvent was dried out in a few minutes. The resulting VCD spectra (with mild baseline correction) is shown in Figure 2.

thumbnail image

Figure 2. VCD spectra of R-camphor film deposited on ICL NaCl IR plate. Spectra taken on JASCO FVS-4000 at the University of Brescia with 1000 accumulation.

Download figure to PowerPoint

Diffuse Reflectance (DR)

Diffuse reflectance (DR) is one of the oldest sampling techniques in the Vis range; since 193544 color analysis has been performed using this approach. This is a potential alternative to the pellet techniques with the advantage of easier sample preparation since a simple mixture between the grinded sample and the scattering powder (or occasionally using only properly powdered sample) is enough. When solid–solid chemical interaction between sample and KBr powder may be a problem,45 Teflon powder is a suitable alternative for its good scattering properties and inert nature.

The measuring principle is simple: circularly polarized radiation enters through an aperture of an integrating sphere, internally coated with BaSO4, MgO or made of Teflon, and hits the sample placed at an aperture on the opposite side. The diffusely reflected radiation is collected and reflected many times by the inside surface of the sphere and at last reaches the photomultiplier tube placed on a third aperture. In an ideal sphere the area of the three apertures is negligible with respect to the total area, to allow the largest possible number of reflections: this means that a good sphere is not so efficient since most of the radiation is absorbed by the inner sphere surface. This limits the signal-to-noise performances particularly in the low UV region; several DR applications in the IR are possible and have been pursued (Gold coated spheres) but no DR VCD has been so far reported, due to the expected sensitivity problems.

Sample dilution with a transparent powder is necessary to reduce spurious influence of the specularly reflected component, proper grinding (as done for preparing Nujol mulls or pellets) is essential here too for the same reasons. Nevertheless qualitative data can occasionally be obtained without dilution and even with no grinding, but related spectra must be treated with much care.

Here too the best results come from very diluted mixtures since AF is present; in addition the linearity range is significantly smaller than in the transmission sampling methods.

Due to the above limitations and the fact that no regular commercial accessory is yet available, reported applications are still very few,46–52 but the easy way to get spectra from all sorts of materials may encourage the use of the technique in the future. In Figure 3 we compare Nujol mull and DR spectra of a Rhodium chiral complex.

thumbnail image

Figure 3. Nujol mull transmission (black) and DR (gray) CD spectra of Λ-Rh[(S,S)Phbdtp]3 complex collected on the J-810 of the University of Bologna, with 1 and 6 nm SBW respectively, 1 sec response, 100 nm/min scanning speed.

Download figure to PowerPoint

In addition to DR, other reflectance type measurements, widely employed in IR spectroscopy could be used. A few years ago an attempt to measure CD in external reflectance mode, at an angle close to grazing incidence, with an homemade experimental accessory, was reported,53 spectra included a lot of artifacts coming from linearly polarized components, but still useful information was extracted.

Attenuated total reflectance (ATR) is a widely used sampling technique particularly in the IR region.54 No CD data have been reported so far; just a simple feasibility experiment has been carried out with a JASCO J-815 spectrometer inserting a right angle quartz prism in the sample compartment and placing the photomultiplier tube on the built-in 90° port. The incoming beam is internally reflected on the hypotenuse of the ATR crystal, where an evanescent wave extends beyond the prism surface. If a sample is placed in close contact to the hypotenuse surface it partially absorbs the evanescent radiation, while most of the light is reflected. To carry on the experiment, the aqueous solutions of the two enantiomers of Λ-Co(en)3Cl3 complex (kind gift of Prof. Sven E. Harnung of University of Copenhagen) were dried on the prism hypotenuse obtaining rather irregular films and spectra have been scanned (Fig. 4). Considering the experimental limitations (very poor quality of the sample layer, approximate alignment of the prism in the compartment) and the 45° incidence angle used which for Total Reflectance will be mainly measured, the results are not so bad. Possible reasons of the different intensities are attributed to AF, as estimated and measured in transmission on the Nujol mull of the same compound.

thumbnail image

Figure 4. Reflectance-CD and absorption spectra of the Λ and Δ enantiomers of Co(en)3Cl3 complex in black and gray respectively. Spectra taken at JASCO Europe, Cremella, Italy on a modified geometry of the J-815 apparatus. 2 nm SBW, 0.5 sec response, 100 nm/min scanning speed.

Download figure to PowerPoint

Liquid Suspensions

Another form in which solid samples may be analyzed is suspensions in a liquid, which means either insoluble chiral powders or chiral molecules bound to insoluble particles, aggregates etc. This topic is of wide interest today since it finds applications in the fields of biocatalysis or even therapeutics as illustrated in the literature cited below.

After an early experimental work54 with a manual/nonscanning spectrometer, an interesting application appeared using a liquid medium with the refractive index matching that of the powder.55

The CD in the low UV of immobilized enzymes on silica gel particles (tens to hundreds of microns large) has been the subject of a well detailed article,56 with the aim to obtain secondary structure estimates, with use of short path cells close to the photomultiplier tube surface, to minimize scattering effects. Sample sedimentation had to be avoided: this was achieved with a motor driven cylindrical cell holder, continuously rotating the cell around the light propagation axis. The AF effects were corrected on an empirical basis, but the cited article stimulated a work of ours57 and another article58 from one of the coworkers of the original article on the AF subject.

When the particles have a still smaller diameter, as in case of nanoparticles, experimental difficulties are much less.59–62

As a caution: conventional stirring of a suspension in a cuvette may create vortices with induced supramolecular optical activity as pointed out recently.63–66

Another recent application deals with nanocrystalline benzophenone67 where CD spectra in Nujol mull and in suspension are compared, the experimental part is however not well described.

Films

The last, and sometimes more challenging, solid sampling system is as films. The term film is rather generic and includes very different types from stand alone thin sheets, to solutions deposited and dried on a transparent window, to Langmuir-Blodgett (LB) depositions, to thin layers between quartz windows .…

Depending on the preparation process, the chiral sample may be more or less oriented and potential artifacts due to anisotropic structure may be present in the recorded CD spectra.

While cautions on interpretation of CD of aligned film had been reported early,68 the problem was properly addressed by a article by Shindo in Biopolymers,69 criticizing a previous article on the same journal70 about the induced optical activity by Congo red.

The film technique has particular relevance for CD of biomolecules; in the IR region the obvious advantage is removal of the overlapping water absorption in the amide I region. After early studies on polypeptides using the dispersive type VCD apparatus in Chicago,71–73 in recent years consistent progress has been reported by the group of Polavarapu,74–76 where alternative types of film preparation have been proposed (supported films, free-standing films, film-on-film, matrix assisted film formation) with remarkable results.75 An additional approach, the “Spray-Method”, had been proposed by the Nafie's group.77

In the UV range after the pioneering work of Fasman78 and Tunis-Schneider and Maestre,79 Brahms et al. studied DNA films80 with an home-made spectrometer checking the absence of linear dichroism parasitic effects. Cassim et al. reported CD on films of purple membranes,81, 82 photoreceptor membranes. A process to create multilayer films (isopotential spin-dry method) has been studied by Clark et al.83, 84 and the method was used by Bazzi and Woody.85

More recently α-helix-bundle proteins have been studied as LB films (and in solution).86

Surface pressure effects in LB protein films were studied by CD and other techniques by Nicolini et al.87

A tendency to increase β-sheet structure in proteins' films dried on a thin quartz window compared to solution phase samples was observed by Hu et al.,88 whose results were however criticized by Harada and Kuroda89 using the dedicated J-800KCM instrument.10 Nakagawa et al. studied amino acid films prepared with vacuum sublimation technique.90 A far more detailed article on CD of poly-methyl-glutamate LB films deposited on a quartz window, with care to avoid linear orientation artifacts and employing long accumulation times to be able to obtain the weak signal, was published by Nordén group.91 Ultra-thin protein films on gold surfaces implied that the averaging of 999 spectra was needed to get a CD signal of acceptable signal to noise.92 As an alternative, to increase the CD signal, a stack of quartz slides with deposited LB films of peptides had been used.93

Oriented circular dichroism (OCD) as introduced by Huang et al.94–97 is another sort of solid-state sampling technique: membranes containing peptides/proteins are stacked to form a multilayer system and spectra are collected under controlled humidity and temperature conditions. Other research groups are involved in this tematic.98, 99

A huge amount of literature data is available on CD film analysis and is often related to material sciences.

Thermal helix inversion in the solid thin films of helical polymers has been reported for poly(L-aspartic acid ester),100 poly(thiophene),101, 102 poly(isocyanate),103 poly (silane)104 and poly(phenylacetylene) spin-casted films by solvent vapor and heating.105, 106

Treated films of fluorene polymers show very strong CD effects, spin coated films of different thickness had been showing a strong nonlinearity between absorption and CD.107 Linear birefringence and dichroism were the probable reason of the phenomena, but no specific measurement had been carried out.

Photoinduced chirality on azomaterial films deposited on quartz or BaF2 windows (by irradiation with circularly polarized Ar+ laser source) was recently measured in both VIS and IR range108; the literature on this subject is very large due to the potential industrial interest on these materials.109–111

Supramolecular assemblies J-aggregates in PVA/aqueous solution were dried as a film and mounted between optical glass plates112: CD spectra were collected rotating the sample around the optical axis and about the vertical axis. More recently Yao et al.113 revisited the same topics in solution phase detecting different conformers according to the apparent CD measurements.

The apparent CD spectra had been used also by Shindo114 to discriminate between achiral polymeric films with one or with multiple layer structure, simply collecting the CD spectra from front and back sides.

The spectral range in which film techniques are widely used is the vacuum ultraviolet region, to remove interferences from solvent absorption. Many applications have been reported using conventional dispersive,115, 116 synchrotron radiation117, 118 or polarizing undulator119, 120 based CD spectrometers.

Film techniques may be subject to scattering effects too: Figure 5 shows spectra of dairy products (low fat yogurt and butter) measured as a film between two quartz plates in the UV region. Both these samples have a very strong intrinsic scattering. Our spectra were obtained placing the sample in the normal cell position of the cell holder and, in a second run, rather close to the photomultiplier tube sensitive surfaces. The data clearly show that, while the apparent absorbance spectra are obviously much less influenced by scattering in the close by position, the apparent shapes of the CD spectra are substantially kept, despite the strong increase in the noise.

thumbnail image

Figure 5. CD (top) and UV (bottom) spectra of commercial low fat yogurt (left) and butter samples. Films obtained by squeezing the product between two quartz plates. Spectra obtained at JASCO Europe, Cremella, Italy, on a J-815 apparatus. 2 nm SBW, 1 sec response, 100 nm/min scanning speed. Gray curves refer to the films placed in the normal sampling position, black ones putting the sample as close as possible to the photomultiplier tube.

Download figure to PowerPoint

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION OF THE LITERATURE AND RESULTS
  5. CONCLUSIONS
  6. LITERATURE CITED
  7. APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

We have reviewed here the existing literature dealing with CD measurements of solid state samples in the form of single crystals, Mulls, Pellets, films on IR plates or else, as Suspensions or measured with DR or ATR/Total Reflectance techniques. We report examples from our own labs of Nujol Mulls CD spectra, of VCD of natural products on IR plates, of a DR CD spectrum on a Rh-complex, of ATR/Total Reflectance-CD on a sample used as a standard and of a CD spectrum of a food-sample squeezed between plates.

To summarize, good practice in getting transmission CD spectra of solid samples should include the following steps:

  • always collect the absorption spectra simultaneously with CD spectra;

  • carry on (if possible) measurements at different concentrations/thickness;

  • measure also linear dichroism spectra (this is easy on VCD units using commercial PEM and lock-in amplifier; for ECD spectrometers an optional accessory is necessary, but this allows, with simple modifications, simultaneous CD and LD collection);

  • place the sample close to photomultiplier tube sensitive surface;

  • carry on measurements by rotating the sample around the light propagation axis at different angles;

  • measure spectra from front and back sides.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION OF THE LITERATURE AND RESULTS
  5. CONCLUSIONS
  6. LITERATURE CITED
  7. APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. DISCUSSION OF THE LITERATURE AND RESULTS
  5. CONCLUSIONS
  6. LITERATURE CITED
  7. APPENDIX: ARE THERE WAYS TO CORRECT LINEAR ANISOTROPIES?

It's well-known that solid state CD spectra measured with commercial spectrometers using polarization-modulation technique may include artifacts originating from the interaction between macroscopic sample anisotropies [linear dichroism (LD) and linear birefringence (LB)] coupled with the non ideal optical properties of the components used in the current CD spectrometers, which use photoelastic modulators for phase modulation spectroscopy. Several articles have dealt with these aspects using Jones or Stokes-Mueller formalisms.121–124

The observed CD signal depends not only on the real CD of the sample, but also on LD and LB, the latter components are often order of magnitudes larger than the true CD signal to be measured.

The measured CD signal can be expressed in a simplified form as Shindo123 and Kuroda6 report:

  • equation image

where, CDapp is CD measured by the instrument, G0 is an apparatus constant, CDtrue is the real CD of the sample, α is the residual static birefringence of modulator, θ is the angular rotation of the sample, LD is linear dichroism and LD′ is 45° linear dichroism, LB is linear birefringence and LB′ is 45° linear birefringence.

The attempt to collect several measurements by rotating the sample about the light propagation axis by an angle θ is able to detect the presence rather than to compensate the linear anisotropies effects. The most advanced approach of Shindo10 calls for the careful choice of specifically selected components for the conventional optical layout (to minimize α) and for the use of an analyzer to set θ at an angle where LB = 0. The true CD signal can then be extracted averaging the two measurements obtained rotating the sample 180° with respect to the vertical axis.

Two recent articles on this subject,125, 126 while related to CD in the IR (VCD), make use of other similar simplified approaches. Once again in the VCD field Laurence Nafie proposed intriguing hardware modifications,127, 128 but still in line with the previously mentioned ideas (Nafie, personal communication). Similar effects of linear anisotropies have also been considered in VUV polarizing undulator.129 In any case a practical way to fully compensate large linear anisotropies does not exist with conventional spectrometers. The best approach requires at least the measurement not only of CD, but also of LD since similar shapes between the two spectra are likely to indicate the presence of artifacts in the CD spectra. Collection of both face and back sample orientation spectra is an additional, simple to achieve, must.