- Email: [email protected]
Crystalline chlorophyll and bacteriochlorophyll
Crystalli.ne Chlorophyll and Bacteriochlorophyll Earl E. Jacobs, Albert E. Vatter and A. Stanley Holt From
the Department
of Botany, University
of
Illinois,
Urbana,
Illinois
Received March 29, 1954
This paper presents evidence that chlorophyll and bacterioo-orchl phyll can be crystallized. Solid chlorophyll usually appears as a “wax.” Willstatter and Stoll (1) and Fischer (2) described chlorophyll precipitated by evaporation from ether-petroleum ether as a “microcrystalline” powder, consisting of thin, lancet-shaped plates; but Hanson’s (3) x-ray study of such preparations showedno sharp diffraction pattern, and since such a pattern is the ultimate criterion of crystallinity, it has been generally accepted that solid chlorophyll is amorphous. It has been suggested that the long phytyl side chain prevents the chlorophyll molecules from forming an orderly array. If the phytyl group is exchanged for an ethyl or methyl group by the action of the enzyme, chlorophyllase, readily crystallizable ethyl or methyl chlorophyllides are obtained. These derivatives have often been called “crystalline chlorophylls,” a misleading nomenclature, which the proof of crystallizability of chlorophyll, to be given below, makes particularly inappropriate. A starting point for renewed study of the crystallization of chlorophyll was provided by contradictory observations concerning the chemical stability of chlorophyll in the dry state. In general, attempts to prepare stable, dry chlorophyll preparations have failed (4, 5) ; but occasional successeshave been reported, without the conditions essential for stability being identified (6-8). When we compared the methods used in these studies, we noticed that relatively stable preparations resulted when the pigment was precipitated from methanol-petroleum ether by washing out the methanol with water (6, S), while samples which deteriorated rapidly were obtained by evaporation or precipitation from anhydrous solvent (4, 5). In light of observations to be described below, 228
CHLOROPHYLL
AND BACTERIOCHLOROPHYLL
229
we believe that the former preparations may have been at least partially crystalline, while the latter ones were amorphous. Convincing indications of chlorophyll crystallization came from spectroscopic measurements on colloidal suspensions.Jacobs et al. (9, 10) found that the red absorption peak of ethyl chlorophyllide is shifted by 80 mp toward longer waves upon crystallization, and that under certain conditions chlorophyll exhibits an equally strong shift (11). Strain (12) had noted a similar, if smaller, band shift-by about 50 mp--when chlorophyll was precipitat,ed by washing-out methanol from petroleum ether-methanol solution with water. It seemedlegitimate to assumethat the extensive shift of the red band indicates the formation of microcrystals, in the caseof chlorophyll as well as in that of chlorophyllide. PREPARATION
AND PROPERTIES
OF CRYSTALS
Preparation of Pure Pigments Certain precautions,suggestedin the earlier literature, proved excessive;for example,variationsin temperature,lighting, andhumidity did not affect the product. Addition of carbonateto prevent pheophytin formation proved unnecessary whenspinachwasusedassourceof chlorphyll. Chlorophylls a and b Eight poundsof fresh spinachwereblendedin 8 1. of acetone; pigmentswere transferred to 1 1. of petroleumether (Skelly Solvent F) by addition of distilled water. The petroleumether layer waswashedsuccessivelywith eight 11. portions of So% aqueous methanol (reagent-grade methanol) to remove carotenols. (Most of the oarotenols precipitated after two washings and were removed immediately by filtration.) The solution was then washed with water until chlorophyll precipitated. The pigment was collected on diatomaoeous earth, washed with 4 1. of petroleum ether (to remove carotenoids), and redissolved in acetone. After transfer to 500 ml. of ethyl ether (Mallinkrodt Analytical Reagent), acetone was removed by exhaustive washing with distilled water, and the solution was dried under vac-
uum. Chlorophylls a and b were separated as follows : The dried sample was dissolved in 10 ml. of pyridine, and the solution was diluted to 1 1. with petroleum ether. The pigments were adsorbed on a powdered sucrose column and separated by 0.5% isopropyl alcohol in pentane (Phillips “pure grade”). Chlorophyll a, which moved first, was collected in the suction flask. Five hundred milliliters of ethyl ether were added to this solution, which was then washed thoroughly with water, leaving a water-saturated, ether-pentane solution of chlorophyll a. Chlorophyll b, remaining on the sugar, was obtained as follows: Sugar above the main band was removed, and ether passed through the column until it reached the bottom of the band. The remaining sugar column was then removed intact and sliced. The pigment-bearing slice was immediately extracted by reagent-grade
230
JACOBS,
VATTER
AND
HOLT
acetone and transferred to 500 ml. of ether by the addition of water. Acetone was removed by exhaustive washing with distilled water, leaving a water-saturated ethereal solution of chlorophyll b.
Bacteriochlorophyll Approximately 200 g. (wet weight) of Rhodospirillum rubrum were extracted with 11. of methanol. Two liters of acetone and 2 1. of ether were added, and the pigment was transferred to ether by washing with water saturated with sodium chloride. Two liters of petroleum ether were added, and the pigment was precipitated by removing the ethyl ether under vacuum. The precipitate was collected by filtration through diatomaceous earth and was washed with 4 1. of petroleum ether to remove most of the accessory pigments. The pigment was redissolved in acetone and transferred to ether by washing with water, and the moist ether solution was evaporated to dryness. Details of chromatographic procedure are described by Holt and Jacobs (13). The pigment was removed from the sugar by acetone and transferred to 500 ml. ether. The solution was washed with water to remove acetone, leaving bacteriochlorophyll in ether saturated with water.
Preparation of Crystals Successful crystallization of the chlorophyll pigments depends on high purity and the presence of water. If these requirements are satisfied, several procedures can be used to obtain crystalline preparations from solutions prepared as described above. They include (a) addition of distilled water forming a layer under the ether solution, and removal of ether under vacuum; (6) addition of an equal volume of hexane (Phillip’s “pure grade”), evaporation of sufhcient ether to cause abundant formation of microcrystals, addition of water to form a layer under the organic solvents, and removal of the latter by evaporation under vacuum; and (c) addition of pentane, repeated washing of the ether-pentane solution with water (in which ether is about 200 times more soluble than pentane) until microcrystals are formed, addition of a layer of water, and removal of the organic solvents under vacuum. If the final evaporation is done carefully enough, a single flake of pigment, which can easily be removed and dried under vacuum, is formed on the water surface. Before these three methods were developed we used another, less satisfactory one, which consisted in adding calcium chloride (cu. 100 p.p.m. of calcium proved sufficient) to a colloidal suspension, obtained by diluting an acetonic solution of chlorophyll with water. Analysis showed that calcium is not incorporated into the crystals; apparently, it neutralizes the charge on the colloidal droplets, permitting them to coalesce and form crystals. (In the absence of calcium, colloidal chlorophyll particles appear under the electron microscope as small spheres and their spectrum shows a peak at 670-675 rnp, characteristic of amorphous pigment.) The importance of water for the crystallization of chlorophylls a and b could be demonstrated as follows: Crystalline pigment was dried under high vacuum for 24 hr. and redissolved in anhydrous ether. The solution was divided into four parts, to one of which was added water, to another an equal volume of hexane, and to a third an equal volume of hexane plus water; the fourth was kept as a control.
CHLOROPHYLL
AND
231
BACTERIOCHLOROPHYLL
Ether and hexane were then removed under vacuum. Crystals (as indicated by sharp x-ray diffraction patterns) were found only in the two samples to which water had been added. Contrary to our preliminary announcement (14), traces of water were found to be necessary also for the crystallization of bacteriochlorophyll. Our experiments indicated that traces of water are needed also for the crystallization of alkpl chlorophyllides, although these compounds crystallize much more easily than the chlorophylls. Pheophytins, on the other hand, crystallized (e.g., by evaporation from ethereal solutions) also under the most strictly anhydrous conditions we were able to realize. The need for water thus seems to be related to the presence of the magnesium atom. The difficulty of chlorophyll crystallization may be related to the combination, in the same molecule, of a long nonpolar chain (phytol) and a rather strong polar atom (magnesium) ; in the ethyl chlorophyllides, the hydrophobic hydrocarbon chain is short; while in pheophytins, the hydrophilic magnesium atom is absent; both crystallize easily-one in the presence, the other also in the absence of water.
Properties
of Crystals
X-ray diffraction patterns indicating the crystalline structure of the samples prepared by the above-described methods are shown in Fig. 1, together with that of noncrystalline chlorophyll, as obtained by Hanson (3). To show that the red absorption peak shifts on crystallization, samples of solid chlorophyll were shaken vigorously in pentane to disperse the microcrystals. Extra-heavy liquid paraffin (Standard Oil Co.) was then added, and pent,ane was removed under vacuum. Transmission and scattering measurements were made with a Beckman model B spectrophotometer.’ The absorption curve (corrected for scattering) of the microcrystals of chlorophyll a is shown in Fig. 2. No fluorescence could be noted in these crystals. The infrared absorption band of bacteriochZorophyl1 also shifted to longer wavelengths when the pigment was crystallized. After correction for scattering, the peak is found to lie close to 865 mF. (Uncorrected data show transmission minima at wavelengths up to 890 mp, depending on the size of the crystals.) These results can perhaps explain the observations of Krasnovsky et al. (15) who found that solid films of bacteriochlorophyll show two, and sometimes three, absorption maxima, at 800,850-865 and 890 m#, respectively. It can be suggested that the peak at 800 mp belonged to an amorphous phase, and that at 850465 rnM to a phase in which the crystalline order extended only over a few molecular diameters.The third peak, which was observed only occasionally, and only when urea was added to the solution, may have belonged to larger microcrystals (whose formation could have been induced by urea), and be shifted to 890 rnp by scattering. Microcrystals of chlorophyll, dispersed by vigorous shaking of samples in hexane, were examined under an electron microscope. A preparation of crystalline 1 The measuring technique and the calculation of “true” by eliminating scattering are described in a pa.per on the ethyl chlorophyllide crystjals (in preparation).
absorption absorption
spectrum spectra of
JACOBS,
VATTER
AND
HOLT
F 'IG. 1. X-ray diffraction patterns of samples of chlorophyll a, chlorophyll b, and bacteriochlorophyll. The upper patterns resulted from precipitation of t #he of water. The lower pattern is that obtained by Hans on pkn nent in the presence preparation. The numbers refer to interplanar spacings. (3) 13y anhydrous
CHLOROPHYLL
I, 1.00 -
I I,
AND BACTERIOCHLOROPHYLL
I I I I,
I1
I I,
233 I I I
1
WAVE LENGTH (M,u) FIG. 2. Absorption (solid
spectra curve)
of chlorophyll and chlorophyll
a crystals suspended a in acetone (dashes).
in
mineral
oil
chlorophyll a is shown in Fig. 3. It shows thin crystalline sheets, most of which were rolled into rolls, probably during the drying. The tendency of chlorophyll to crystallize in thin layers is evident from this figure. From the length of the shadow on metal-shadowed preparations, the thickness of the sheets was estimat,ed to be approximately 50 A. corresponding to a few-perhaps only two-molecular layers of chloropyll. The electron photomicrographs of the microcrystals of ethyl chlorophyllide (10) also show a layered structure, but the crystals show a much greater capacity to grow in the third dimension, perpendicular to the sheet. These observations indicate that intermolecular forces in chlorophyll crystals are much stronger in the planar networks than between them. If the red shift of absorption bands in the crystals is due to resonance forces between molecules (as we assume it to be), then this shift must be attributed mainly to interaction within the planes. To test this conclusion, we examined the absorption spectra of surface layers of chlorophyll.
CHLOROPHYLL
SURFACE LAYERS
Hanson (3), Langmuir and Schaefer (16), and Alexander (17) described the preparation of chlorophyll monolayers. They found them to behave, upon compression, as a two-dimensional gas, without evidence of condensation. When the layer was compressedto a density at which further decreasein area required a sharp increase in pressure, the chlorophyll molecules occupied 106 sq. A. each-about 37 sq. A. more than ethyl chlorophyllide molecules occupy in “crystalline” monolayers (Hanson (3) confirmed by our measurments). This extra surface requirement can be attributed to phytol. Such “amorphous” chlorophyll monolayers
234
JACOBS,
FIG.
3. Electron
microscope
VATTER
photographs
AND
HOLT
of chlorophyll
a crystals.
could be prepared by spreading drops (0.005 ml.) of chlorophyll solution in hexane + 1% pyridine (optical density at red peak, 150/cm.) on a clean water surface confined to an area about 100 sq. cm. The absorption spectrum of the films, determined by picking them up on a glass cover slip and stacking several cover slips in the path of the light beam, chlorophyll colloids or adsorbates, was similar to that of “amorphous”
CHLOROPHYLL
AND
235
BACTERIOCHLOROPHYLL
with a red absorption peak at 675 ml.c. The optical density in the peak of the red band was 0.011 per monolayer, in approximate agreement with the value (0.014) calculated from the molar absorption coefficient of chlorophyll a in solution, for a surface density of one molecule per 106 sq. A. The calculation presupposes that the orientation of the molecules in the adsorbed layer does not affect their capacity for absorption of light falling normally to the layer. The approximate agreement between the observed and calculated optical density suggests that this supposition s permissible, and also that the density of the layer is not much changed by transfer from water to the glass surface. In analogy to our preparation of crystalline solid chlorophyll, chlorophyll surface films, whose absorption spectrum indicated crystalline structure, also could be obtained. Droplets of chlorophyll solution on water containing calcium ions showed sharp contraction of the area covered by the spreading droplet; formation of a rigid film was demonstrated by blowing talc across it. In Fig. 4, the absorption spectrum of such surface layers, picked up on glass, is seen to be quite similar to that of suspensions of crystalline chlorophyll. The red absorption peak is at 735 rnp; the optical density per layer in this peak is as high as 0.026, 40% higher than in a monolayer of ethyl chlorophyllide. Hanson (3)
0.005
0.000
400
450
500
550 600 WAVELENGTH
650 IN MU
700
750
FIG. 4. Absorption spectrum of crystalline chlorophyll a film compared with absorption spectrum of noncrystalline chlorophyll circles). Arrows indicate location of absorption maxima in acetone.
(open a film
circles) (closed
JACOBS, VATTER
236
AND HOLT
found that the disk-shaped molecules of ethyl chlorophyllide are arranged, in a surface monolayer (as well as in a three-dimensional crystal) under a 55” angle to the plane of the layer. It seemsimplausible that the surface area requirement of a chlorophyll molecule can be reduced, by a different spatial orientation within the film (such as standing “on edge” at an angle of 90” to the surface), by as much as 40 % below that of ethyl chlorophyllide. Increased capacity for the absorption of light passing normally through. the layer is also unlikely, since in Hanson’s model of ethyl chlorophyllide films, the short axis of the porphin resonance system is parallel to the water layer; according to LonguetHiggins el al. (18), the red absorption band corresponds to vibrations along this axis. We are therefore inclined (19) to ascribe the high optical density of the “crystalline” surface layers of chlorophyll to a bimolecular, rather than unimolecular, structure, perhaps, with phytol chains filling the spacebetween two porphin layers. However, we are not quite certain whether our crystalline films were truly uniform; they may have consisted of a crystal mosaic with an average thickness of about two molecules. STABILITY
OF THE CRYSTALLINE PREPARATIONS
CHLOROPHYLL
As first established by Zscheile and Comar (4), the ratios of the extinction coefficients of chlorophyll in the peaks of the absorption bands and in the minima between them, are sensitive indicators of pheophytin contamination. They are often referred to as “purity ratios,” and their constancy can serve as test of stability in storage, at least in as far as conversion to pheophytin is concerned. Table I shows the purity ratios of our preparations. They are as high as in the purest preparations by Harris and Zscheile (8), and are not noticeably changed in crystalline preparations after 6 months of storage in darkness at room temperature. All samplesgave positive phase tests at the end of this time. In the case of bacteriochlorophyll, the spectroscopic criteria of purity suggestedby Holt and Jacobs (13) were used, i.e., the ratios of the extinctions in the four absorption peaks and in the minimum at 500 rnp. These ratios too, were unchanged by 6 months storage in the crystalline state. The problem of chlorophyll storage seemsto be solved by these findings. Pure compounds are best preserved in crystalline form, which gives protection from the surrounding medium to most molecules and leaves exposed to chemical attack only the relatively few molecules on the crys-
CHLOROPHYLL
AND BACTERIOCHLOROPHYLL TABLE
“Purity
Ratios” of Crystalline before and after 6 Months’
Chlorophyll tone
a in ace-
Before After Chlorophyll tone Before After Bacteriochlorophyll in ether Before After
237
I
Chlorophylls a and b, and Bacteriochlorophyll Storage in Darkness at Room TemperatuTe
-ff.302
-alao
aSO6
cY606
43.6 43.6
54.9 54.7
49.0 49.4
15.9 16.0
ff867 3 a-600 88.0 87.8
-a002 as00 58.2 58.1
b in ace-
-as76 aso0 25.6 25.2
-a772 cc600 116 116
tal surfaces. If these surface molecules do not have any strongly reactive groups exposed, the stability of the preparation is further improved. Highly purified substancesusually crystallize spontaneously when their solubility has been exceeded, and the importance of crystallization for their stability therefore goes unnoticed. In the case of chlorophyll, the difficulty of preparing crystalline Sampleslays emphasis on the importance of crystallization. We now see that it was this difficulty which in the past has frustrated workers trying to prepare chlorophyll material suitable for storage. ACKNOWLEDGMENTS This work was carried out at the Photosynthesis Research Laboratory, Department of Botany, University of Illinois, with the aid of contract NR 119-229 between the Office of Naval Research and University of Illinois. Our thanks are due to Dr. Rabinowitch for advice and encouragement, and to Dr. Emerson and other members of the Photosynthesis Laboratory for friendly cooperation.
SUMMARY
Simple directions are given for the preparation of gram batches of pure crystalline chlorophyll a and b and bacteriochlorophyll. X-ray diffraction patterns of the samplesprove their crystalline nature. Presence
238
JACOBS,
VATTER
AND
HOLT
of water in the crystallizing solvent is the most important factor in crystallization. A shift of the red absorption band toward longer wavelengths by about 80 and 90 rnp accompanies the crystallization of chlorophyll and bacteriochlorophyll, respectively. An analogous shift was previously described for crystalline alkyl chlorophyllides and bacteriochlorophyllides. Spectroscopic measurements show that the crystalline pigment preparations are unchanged after 6 months’ storage at room temperature. Two types of chlorophyll surface layers have been observed, of “compressed gas” and ‘Lcrystalline” type, respectively, showing, when picked up on glass, absorption bands at 675 and 735 rnp, respectively. REFERENCES 1. WILLSTXTTER, R., AND STOLL, A., “Untersuchungen iiber Chlorophyll.” Julius Springer, Berlin, 1913. 2. FISCHER, H., AND STERN, A., “Chemie des Pyrrols,” Vol. II. Akad. Verlag, Leipzig, 1940. 3. HANSON, E. A., Rec. t+av. botan. n8erZ. 36, 183 (1939). 4. ZSCHEILE, F. P., AND COMAR, C. L., Botan. Gaz. 102, 463 (1941). 5. ZSCHEILE, F. P., COMAR, C. L., AND HARRIS, D. G., Plant Physiol. 19, 627 (1944). 6. MACKINNEY, G., J. Biol. Chem. 133, 91 (1940). 7. ZSCHEILE, F. P., COMAR, C. L., AND MACKINNEY, G., Plant Physiol. 17, 666 (1942). 8. HARRIS, D. G., AND ZSCHEILE, F. P., Botan. Gas. 104, 515 (1943). 9. JACOBS, E. E., AND HOLT, A. S., J. Chem. Phys. 20, 1326 (1952). 10. RABINOWITCH, E., JACOBS, E. E., HOLT, A. S., AND KROMHOUT, R., 2. Physik 133, 261 (1952). 11. JACOBS, E. E., AND HOLT, A. S., J. Chem. Phys. 22, 142 (1954). 12. STRAIN, H. H., Science 116, 174 (1952). 13. HOLT, A. S., AND JACOBS, E. E., Am. J. Botany, in press. 14. JACOBS, E. E., VATTER, A. W., AND HOLT, A. S., J. Chem. Phys. al,2246 (1953). 15. KRASNOVSKY, A. A., VOJNOVSKAJA, K. K., AND KOSOBUTSKAJA, L. M., Doklady Akad. Nauk S.S.S.R. 86, 389 (1952). 16. LANGMUIR, I., AND SCHAEFER, V. J., J. Am. Chem. Sot. 69, 2075 (1937). 17. ALEXANDER, A. E., J. Chem. Sot. 1937, 1813. 18. LONGUET-HIGGINS, H. C., RECTOR, C. W., AND PLATT, J. R., J. Chem. Phys. 18, 1174 (1944). 19. JACOBS, E. E., HOLT, A. S., AND RABINOWTICH, E., J. Chem. Phys. 22, 142 (1954).
the Department
of Botany, University
of
Illinois,
Urbana,
Illinois
Received March 29, 1954
This paper presents evidence that chlorophyll and bacterioo-orchl phyll can be crystallized. Solid chlorophyll usually appears as a “wax.” Willstatter and Stoll (1) and Fischer (2) described chlorophyll precipitated by evaporation from ether-petroleum ether as a “microcrystalline” powder, consisting of thin, lancet-shaped plates; but Hanson’s (3) x-ray study of such preparations showedno sharp diffraction pattern, and since such a pattern is the ultimate criterion of crystallinity, it has been generally accepted that solid chlorophyll is amorphous. It has been suggested that the long phytyl side chain prevents the chlorophyll molecules from forming an orderly array. If the phytyl group is exchanged for an ethyl or methyl group by the action of the enzyme, chlorophyllase, readily crystallizable ethyl or methyl chlorophyllides are obtained. These derivatives have often been called “crystalline chlorophylls,” a misleading nomenclature, which the proof of crystallizability of chlorophyll, to be given below, makes particularly inappropriate. A starting point for renewed study of the crystallization of chlorophyll was provided by contradictory observations concerning the chemical stability of chlorophyll in the dry state. In general, attempts to prepare stable, dry chlorophyll preparations have failed (4, 5) ; but occasional successeshave been reported, without the conditions essential for stability being identified (6-8). When we compared the methods used in these studies, we noticed that relatively stable preparations resulted when the pigment was precipitated from methanol-petroleum ether by washing out the methanol with water (6, S), while samples which deteriorated rapidly were obtained by evaporation or precipitation from anhydrous solvent (4, 5). In light of observations to be described below, 228
CHLOROPHYLL
AND BACTERIOCHLOROPHYLL
229
we believe that the former preparations may have been at least partially crystalline, while the latter ones were amorphous. Convincing indications of chlorophyll crystallization came from spectroscopic measurements on colloidal suspensions.Jacobs et al. (9, 10) found that the red absorption peak of ethyl chlorophyllide is shifted by 80 mp toward longer waves upon crystallization, and that under certain conditions chlorophyll exhibits an equally strong shift (11). Strain (12) had noted a similar, if smaller, band shift-by about 50 mp--when chlorophyll was precipitat,ed by washing-out methanol from petroleum ether-methanol solution with water. It seemedlegitimate to assumethat the extensive shift of the red band indicates the formation of microcrystals, in the caseof chlorophyll as well as in that of chlorophyllide. PREPARATION
AND PROPERTIES
OF CRYSTALS
Preparation of Pure Pigments Certain precautions,suggestedin the earlier literature, proved excessive;for example,variationsin temperature,lighting, andhumidity did not affect the product. Addition of carbonateto prevent pheophytin formation proved unnecessary whenspinachwasusedassourceof chlorphyll. Chlorophylls a and b Eight poundsof fresh spinachwereblendedin 8 1. of acetone; pigmentswere transferred to 1 1. of petroleumether (Skelly Solvent F) by addition of distilled water. The petroleumether layer waswashedsuccessivelywith eight 11. portions of So% aqueous methanol (reagent-grade methanol) to remove carotenols. (Most of the oarotenols precipitated after two washings and were removed immediately by filtration.) The solution was then washed with water until chlorophyll precipitated. The pigment was collected on diatomaoeous earth, washed with 4 1. of petroleum ether (to remove carotenoids), and redissolved in acetone. After transfer to 500 ml. of ethyl ether (Mallinkrodt Analytical Reagent), acetone was removed by exhaustive washing with distilled water, and the solution was dried under vac-
uum. Chlorophylls a and b were separated as follows : The dried sample was dissolved in 10 ml. of pyridine, and the solution was diluted to 1 1. with petroleum ether. The pigments were adsorbed on a powdered sucrose column and separated by 0.5% isopropyl alcohol in pentane (Phillips “pure grade”). Chlorophyll a, which moved first, was collected in the suction flask. Five hundred milliliters of ethyl ether were added to this solution, which was then washed thoroughly with water, leaving a water-saturated, ether-pentane solution of chlorophyll a. Chlorophyll b, remaining on the sugar, was obtained as follows: Sugar above the main band was removed, and ether passed through the column until it reached the bottom of the band. The remaining sugar column was then removed intact and sliced. The pigment-bearing slice was immediately extracted by reagent-grade
230
JACOBS,
VATTER
AND
HOLT
acetone and transferred to 500 ml. of ether by the addition of water. Acetone was removed by exhaustive washing with distilled water, leaving a water-saturated ethereal solution of chlorophyll b.
Bacteriochlorophyll Approximately 200 g. (wet weight) of Rhodospirillum rubrum were extracted with 11. of methanol. Two liters of acetone and 2 1. of ether were added, and the pigment was transferred to ether by washing with water saturated with sodium chloride. Two liters of petroleum ether were added, and the pigment was precipitated by removing the ethyl ether under vacuum. The precipitate was collected by filtration through diatomaceous earth and was washed with 4 1. of petroleum ether to remove most of the accessory pigments. The pigment was redissolved in acetone and transferred to ether by washing with water, and the moist ether solution was evaporated to dryness. Details of chromatographic procedure are described by Holt and Jacobs (13). The pigment was removed from the sugar by acetone and transferred to 500 ml. ether. The solution was washed with water to remove acetone, leaving bacteriochlorophyll in ether saturated with water.
Preparation of Crystals Successful crystallization of the chlorophyll pigments depends on high purity and the presence of water. If these requirements are satisfied, several procedures can be used to obtain crystalline preparations from solutions prepared as described above. They include (a) addition of distilled water forming a layer under the ether solution, and removal of ether under vacuum; (6) addition of an equal volume of hexane (Phillip’s “pure grade”), evaporation of sufhcient ether to cause abundant formation of microcrystals, addition of water to form a layer under the organic solvents, and removal of the latter by evaporation under vacuum; and (c) addition of pentane, repeated washing of the ether-pentane solution with water (in which ether is about 200 times more soluble than pentane) until microcrystals are formed, addition of a layer of water, and removal of the organic solvents under vacuum. If the final evaporation is done carefully enough, a single flake of pigment, which can easily be removed and dried under vacuum, is formed on the water surface. Before these three methods were developed we used another, less satisfactory one, which consisted in adding calcium chloride (cu. 100 p.p.m. of calcium proved sufficient) to a colloidal suspension, obtained by diluting an acetonic solution of chlorophyll with water. Analysis showed that calcium is not incorporated into the crystals; apparently, it neutralizes the charge on the colloidal droplets, permitting them to coalesce and form crystals. (In the absence of calcium, colloidal chlorophyll particles appear under the electron microscope as small spheres and their spectrum shows a peak at 670-675 rnp, characteristic of amorphous pigment.) The importance of water for the crystallization of chlorophylls a and b could be demonstrated as follows: Crystalline pigment was dried under high vacuum for 24 hr. and redissolved in anhydrous ether. The solution was divided into four parts, to one of which was added water, to another an equal volume of hexane, and to a third an equal volume of hexane plus water; the fourth was kept as a control.
CHLOROPHYLL
AND
231
BACTERIOCHLOROPHYLL
Ether and hexane were then removed under vacuum. Crystals (as indicated by sharp x-ray diffraction patterns) were found only in the two samples to which water had been added. Contrary to our preliminary announcement (14), traces of water were found to be necessary also for the crystallization of bacteriochlorophyll. Our experiments indicated that traces of water are needed also for the crystallization of alkpl chlorophyllides, although these compounds crystallize much more easily than the chlorophylls. Pheophytins, on the other hand, crystallized (e.g., by evaporation from ethereal solutions) also under the most strictly anhydrous conditions we were able to realize. The need for water thus seems to be related to the presence of the magnesium atom. The difficulty of chlorophyll crystallization may be related to the combination, in the same molecule, of a long nonpolar chain (phytol) and a rather strong polar atom (magnesium) ; in the ethyl chlorophyllides, the hydrophobic hydrocarbon chain is short; while in pheophytins, the hydrophilic magnesium atom is absent; both crystallize easily-one in the presence, the other also in the absence of water.
Properties
of Crystals
X-ray diffraction patterns indicating the crystalline structure of the samples prepared by the above-described methods are shown in Fig. 1, together with that of noncrystalline chlorophyll, as obtained by Hanson (3). To show that the red absorption peak shifts on crystallization, samples of solid chlorophyll were shaken vigorously in pentane to disperse the microcrystals. Extra-heavy liquid paraffin (Standard Oil Co.) was then added, and pent,ane was removed under vacuum. Transmission and scattering measurements were made with a Beckman model B spectrophotometer.’ The absorption curve (corrected for scattering) of the microcrystals of chlorophyll a is shown in Fig. 2. No fluorescence could be noted in these crystals. The infrared absorption band of bacteriochZorophyl1 also shifted to longer wavelengths when the pigment was crystallized. After correction for scattering, the peak is found to lie close to 865 mF. (Uncorrected data show transmission minima at wavelengths up to 890 mp, depending on the size of the crystals.) These results can perhaps explain the observations of Krasnovsky et al. (15) who found that solid films of bacteriochlorophyll show two, and sometimes three, absorption maxima, at 800,850-865 and 890 m#, respectively. It can be suggested that the peak at 800 mp belonged to an amorphous phase, and that at 850465 rnM to a phase in which the crystalline order extended only over a few molecular diameters.The third peak, which was observed only occasionally, and only when urea was added to the solution, may have belonged to larger microcrystals (whose formation could have been induced by urea), and be shifted to 890 rnp by scattering. Microcrystals of chlorophyll, dispersed by vigorous shaking of samples in hexane, were examined under an electron microscope. A preparation of crystalline 1 The measuring technique and the calculation of “true” by eliminating scattering are described in a pa.per on the ethyl chlorophyllide crystjals (in preparation).
absorption absorption
spectrum spectra of
JACOBS,
VATTER
AND
HOLT
F 'IG. 1. X-ray diffraction patterns of samples of chlorophyll a, chlorophyll b, and bacteriochlorophyll. The upper patterns resulted from precipitation of t #he of water. The lower pattern is that obtained by Hans on pkn nent in the presence preparation. The numbers refer to interplanar spacings. (3) 13y anhydrous
CHLOROPHYLL
I, 1.00 -
I I,
AND BACTERIOCHLOROPHYLL
I I I I,
I1
I I,
233 I I I
1
WAVE LENGTH (M,u) FIG. 2. Absorption (solid
spectra curve)
of chlorophyll and chlorophyll
a crystals suspended a in acetone (dashes).
in
mineral
oil
chlorophyll a is shown in Fig. 3. It shows thin crystalline sheets, most of which were rolled into rolls, probably during the drying. The tendency of chlorophyll to crystallize in thin layers is evident from this figure. From the length of the shadow on metal-shadowed preparations, the thickness of the sheets was estimat,ed to be approximately 50 A. corresponding to a few-perhaps only two-molecular layers of chloropyll. The electron photomicrographs of the microcrystals of ethyl chlorophyllide (10) also show a layered structure, but the crystals show a much greater capacity to grow in the third dimension, perpendicular to the sheet. These observations indicate that intermolecular forces in chlorophyll crystals are much stronger in the planar networks than between them. If the red shift of absorption bands in the crystals is due to resonance forces between molecules (as we assume it to be), then this shift must be attributed mainly to interaction within the planes. To test this conclusion, we examined the absorption spectra of surface layers of chlorophyll.
CHLOROPHYLL
SURFACE LAYERS
Hanson (3), Langmuir and Schaefer (16), and Alexander (17) described the preparation of chlorophyll monolayers. They found them to behave, upon compression, as a two-dimensional gas, without evidence of condensation. When the layer was compressedto a density at which further decreasein area required a sharp increase in pressure, the chlorophyll molecules occupied 106 sq. A. each-about 37 sq. A. more than ethyl chlorophyllide molecules occupy in “crystalline” monolayers (Hanson (3) confirmed by our measurments). This extra surface requirement can be attributed to phytol. Such “amorphous” chlorophyll monolayers
234
JACOBS,
FIG.
3. Electron
microscope
VATTER
photographs
AND
HOLT
of chlorophyll
a crystals.
could be prepared by spreading drops (0.005 ml.) of chlorophyll solution in hexane + 1% pyridine (optical density at red peak, 150/cm.) on a clean water surface confined to an area about 100 sq. cm. The absorption spectrum of the films, determined by picking them up on a glass cover slip and stacking several cover slips in the path of the light beam, chlorophyll colloids or adsorbates, was similar to that of “amorphous”
CHLOROPHYLL
AND
235
BACTERIOCHLOROPHYLL
with a red absorption peak at 675 ml.c. The optical density in the peak of the red band was 0.011 per monolayer, in approximate agreement with the value (0.014) calculated from the molar absorption coefficient of chlorophyll a in solution, for a surface density of one molecule per 106 sq. A. The calculation presupposes that the orientation of the molecules in the adsorbed layer does not affect their capacity for absorption of light falling normally to the layer. The approximate agreement between the observed and calculated optical density suggests that this supposition s permissible, and also that the density of the layer is not much changed by transfer from water to the glass surface. In analogy to our preparation of crystalline solid chlorophyll, chlorophyll surface films, whose absorption spectrum indicated crystalline structure, also could be obtained. Droplets of chlorophyll solution on water containing calcium ions showed sharp contraction of the area covered by the spreading droplet; formation of a rigid film was demonstrated by blowing talc across it. In Fig. 4, the absorption spectrum of such surface layers, picked up on glass, is seen to be quite similar to that of suspensions of crystalline chlorophyll. The red absorption peak is at 735 rnp; the optical density per layer in this peak is as high as 0.026, 40% higher than in a monolayer of ethyl chlorophyllide. Hanson (3)
0.005
0.000
400
450
500
550 600 WAVELENGTH
650 IN MU
700
750
FIG. 4. Absorption spectrum of crystalline chlorophyll a film compared with absorption spectrum of noncrystalline chlorophyll circles). Arrows indicate location of absorption maxima in acetone.
(open a film
circles) (closed
JACOBS, VATTER
236
AND HOLT
found that the disk-shaped molecules of ethyl chlorophyllide are arranged, in a surface monolayer (as well as in a three-dimensional crystal) under a 55” angle to the plane of the layer. It seemsimplausible that the surface area requirement of a chlorophyll molecule can be reduced, by a different spatial orientation within the film (such as standing “on edge” at an angle of 90” to the surface), by as much as 40 % below that of ethyl chlorophyllide. Increased capacity for the absorption of light passing normally through. the layer is also unlikely, since in Hanson’s model of ethyl chlorophyllide films, the short axis of the porphin resonance system is parallel to the water layer; according to LonguetHiggins el al. (18), the red absorption band corresponds to vibrations along this axis. We are therefore inclined (19) to ascribe the high optical density of the “crystalline” surface layers of chlorophyll to a bimolecular, rather than unimolecular, structure, perhaps, with phytol chains filling the spacebetween two porphin layers. However, we are not quite certain whether our crystalline films were truly uniform; they may have consisted of a crystal mosaic with an average thickness of about two molecules. STABILITY
OF THE CRYSTALLINE PREPARATIONS
CHLOROPHYLL
As first established by Zscheile and Comar (4), the ratios of the extinction coefficients of chlorophyll in the peaks of the absorption bands and in the minima between them, are sensitive indicators of pheophytin contamination. They are often referred to as “purity ratios,” and their constancy can serve as test of stability in storage, at least in as far as conversion to pheophytin is concerned. Table I shows the purity ratios of our preparations. They are as high as in the purest preparations by Harris and Zscheile (8), and are not noticeably changed in crystalline preparations after 6 months of storage in darkness at room temperature. All samplesgave positive phase tests at the end of this time. In the case of bacteriochlorophyll, the spectroscopic criteria of purity suggestedby Holt and Jacobs (13) were used, i.e., the ratios of the extinctions in the four absorption peaks and in the minimum at 500 rnp. These ratios too, were unchanged by 6 months storage in the crystalline state. The problem of chlorophyll storage seemsto be solved by these findings. Pure compounds are best preserved in crystalline form, which gives protection from the surrounding medium to most molecules and leaves exposed to chemical attack only the relatively few molecules on the crys-
CHLOROPHYLL
AND BACTERIOCHLOROPHYLL TABLE
“Purity
Ratios” of Crystalline before and after 6 Months’
Chlorophyll tone
a in ace-
Before After Chlorophyll tone Before After Bacteriochlorophyll in ether Before After
237
I
Chlorophylls a and b, and Bacteriochlorophyll Storage in Darkness at Room TemperatuTe
-ff.302
-alao
aSO6
cY606
43.6 43.6
54.9 54.7
49.0 49.4
15.9 16.0
ff867 3 a-600 88.0 87.8
-a002 as00 58.2 58.1
b in ace-
-as76 aso0 25.6 25.2
-a772 cc600 116 116
tal surfaces. If these surface molecules do not have any strongly reactive groups exposed, the stability of the preparation is further improved. Highly purified substancesusually crystallize spontaneously when their solubility has been exceeded, and the importance of crystallization for their stability therefore goes unnoticed. In the case of chlorophyll, the difficulty of preparing crystalline Sampleslays emphasis on the importance of crystallization. We now see that it was this difficulty which in the past has frustrated workers trying to prepare chlorophyll material suitable for storage. ACKNOWLEDGMENTS This work was carried out at the Photosynthesis Research Laboratory, Department of Botany, University of Illinois, with the aid of contract NR 119-229 between the Office of Naval Research and University of Illinois. Our thanks are due to Dr. Rabinowitch for advice and encouragement, and to Dr. Emerson and other members of the Photosynthesis Laboratory for friendly cooperation.
SUMMARY
Simple directions are given for the preparation of gram batches of pure crystalline chlorophyll a and b and bacteriochlorophyll. X-ray diffraction patterns of the samplesprove their crystalline nature. Presence
238
JACOBS,
VATTER
AND
HOLT
of water in the crystallizing solvent is the most important factor in crystallization. A shift of the red absorption band toward longer wavelengths by about 80 and 90 rnp accompanies the crystallization of chlorophyll and bacteriochlorophyll, respectively. An analogous shift was previously described for crystalline alkyl chlorophyllides and bacteriochlorophyllides. Spectroscopic measurements show that the crystalline pigment preparations are unchanged after 6 months’ storage at room temperature. Two types of chlorophyll surface layers have been observed, of “compressed gas” and ‘Lcrystalline” type, respectively, showing, when picked up on glass, absorption bands at 675 and 735 rnp, respectively. REFERENCES 1. WILLSTXTTER, R., AND STOLL, A., “Untersuchungen iiber Chlorophyll.” Julius Springer, Berlin, 1913. 2. FISCHER, H., AND STERN, A., “Chemie des Pyrrols,” Vol. II. Akad. Verlag, Leipzig, 1940. 3. HANSON, E. A., Rec. t+av. botan. n8erZ. 36, 183 (1939). 4. ZSCHEILE, F. P., AND COMAR, C. L., Botan. Gaz. 102, 463 (1941). 5. ZSCHEILE, F. P., COMAR, C. L., AND HARRIS, D. G., Plant Physiol. 19, 627 (1944). 6. MACKINNEY, G., J. Biol. Chem. 133, 91 (1940). 7. ZSCHEILE, F. P., COMAR, C. L., AND MACKINNEY, G., Plant Physiol. 17, 666 (1942). 8. HARRIS, D. G., AND ZSCHEILE, F. P., Botan. Gas. 104, 515 (1943). 9. JACOBS, E. E., AND HOLT, A. S., J. Chem. Phys. 20, 1326 (1952). 10. RABINOWITCH, E., JACOBS, E. E., HOLT, A. S., AND KROMHOUT, R., 2. Physik 133, 261 (1952). 11. JACOBS, E. E., AND HOLT, A. S., J. Chem. Phys. 22, 142 (1954). 12. STRAIN, H. H., Science 116, 174 (1952). 13. HOLT, A. S., AND JACOBS, E. E., Am. J. Botany, in press. 14. JACOBS, E. E., VATTER, A. W., AND HOLT, A. S., J. Chem. Phys. al,2246 (1953). 15. KRASNOVSKY, A. A., VOJNOVSKAJA, K. K., AND KOSOBUTSKAJA, L. M., Doklady Akad. Nauk S.S.S.R. 86, 389 (1952). 16. LANGMUIR, I., AND SCHAEFER, V. J., J. Am. Chem. Sot. 69, 2075 (1937). 17. ALEXANDER, A. E., J. Chem. Sot. 1937, 1813. 18. LONGUET-HIGGINS, H. C., RECTOR, C. W., AND PLATT, J. R., J. Chem. Phys. 18, 1174 (1944). 19. JACOBS, E. E., HOLT, A. S., AND RABINOWTICH, E., J. Chem. Phys. 22, 142 (1954).