Proc. Natl. Acad. Sci. USA Vol. 77, No. 4, pp. 1745-1748, April 1980 Chemistry

Self-assembled a systems as studied by californium-252 plasma desorption mass spectroscopy (oligomer formation/molecular ions/fragment ions/antenna chlorophyll) J. E. HUNT*, R. D. MACFARLANE*, J. J. KATZt, AND R. C. DOUGHERTYt *Department of Chemistry, Texas A & M University, College Station, Texas 77843; tChemistry Division, Argonne National Laboratory, Argonne, Illinois 60439; and IDepartment of Chemistry, Florida State University, Tallahassee, Florida 32306 Contributed by Joseph J. Katz, January 24, 1980

ABSTRACT Self-assembled and H a systems in thin solid films have been studied by 2'52Cf plasma \ 2b desorption mass spectrometry (PDMS). The 252Cf-PDMS spectra of these films show monomer cation and anion molecular ions, H ions of molecular aggregates, and positive and negative ion fragmentation patterns arising from the loss of various aliphatic side chains from the ring. Chlorophyll a films cast from dry carbon tetrachloride solution, in which chlorophyll a is known to occur as the dimer, produced an abundant dimer ion. The highest degree of chlorophyll a self-assembly was observed in chlorophyll a films cast from n-octane solutions. Oligomer ions extending upwards in size to the heptamer were detected in this system. It has long been recognized that the optical properties of chlo- rophyll in vivo are anomalous (1). Relative to a solution of chlorophyll in a polar solvent such as pyridine or diethyl ether, an environment in which chlorophyll is known to exist pre- dominantly as monomer (2), the main red absorption band of chlorophyll in vivo is both red-shifted and broadened. Both the absorption and emission spectra are significantly different (3). C-phy 2 The discrepancies are most often attributed (i) to aggregation of the pigment molecules or (ii) to specific chlorophyll-protein C-phy I C-phy 3 C-phy 7 C-phy 11 interactions. Krasnovskii (4), Brody and Brody (5, 6), and others have suggested that chlorophyll occurs in various states of phy:= CH2 aggregation in vivo, and that interactions between the chloro- phyll molecules in the various aggregates perturb the electronic C-phy 3a states of the and, thus, become the primary FIG. 1. Structure and numbering system for chlorophyll a (Chl source of the spectral anomalies in vivo. Whether there are a). Pheophytin a (Pheo a) is Chl a with 2 H in place of Mg. phy, specific chlorophyll-protein interactions that can generate Phytyl. spectral shifts in the absence of pigment -chro- mophore interactions is still an unresolved question. The nature of the various red-shifted chlorophyll systems that combination of electrophilic Mg and nucleophilic keto C=O can readily be prepared in the laboratory thus becomes relevant group leads to self-aggregation under conditions in which ex- to the state of chlorophyll in vivo. Organic compounds with traneous nucleophiles are absent. large delocalized 7r electron systems (e.g., chlorophyll) expe- We denote a chlorophyll aggregate resulting from coordi- rience nonspecific 7r-7r intermolecular attractive dispersion nation interactions intrinsic to the chlorophyll molecular forces that result in aggregation. In addition, chlorophyll is now structure as "self-assembled." Such systems can be said to be recognized to have coordination properties (7) particularly organized, although they may not have the degree of order conducive to the formation both of chlorophyll-chlorophyll characteristic of micelles, bilayer lipid membranes, or mono- (endogamous) adducts (8), and chlorophyll-nucleophile (exo- layer assemblies (14). Self-assembled chlorophyll systems gamous) aggregates (9). Nuclear magnetic resonance (10, 11) crosslinked by water or ethanol are excellent models for green and infrared (12) spectroscopy show that the central plant photoreaction center chlorophyll (P700) in vivo (15), and atom of chlorophyll with coordination number 4 (as shown in a-water adducts closely mimic the optical Fig. 1) is coordinatively unsaturated. The Mg atom thus has a properties of intact photosynthetic bacteria (16). Chloro- strong tendency to acquire electrons by filling one or both of phyll-chlorophyll oligomers, (Chl a)X, formed by keto C'=O its axial positions with nucleophilic ligands that have lone pairs ... Mg coordination (12, 13) interactions have been suggested of electrons. At the same time, the keto C=O function in ring as a model for green plant antenna chlorophyll (17). V has been shown to be an excellent nucleophile (12, 13). The The composition of self-assembled chlorophyll systems has been difficult to establish because of their large size and lability. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "ad- Abbreviations: Chl a, chlorophyll a; Pheo a, pheophytin a; TOF, vertisement" in accordance with 18 U. S. C. §1734 solely to indicate time-of-flight; PDMS, plasma desorption mass spectroscopy; CEMA, this fact. channel electron multiplier array. 1745 Downloaded by guest on September 26, 2021 1746 Chemistry: Hunt et al. Proc. Natl. Acad. Sci. USA 77 (1980) 252Cf plasma desorption mass spectroscopy (PDMS) is a tech- The TOF data up to 26 bits, binary, were transferred through nique in which ions are produced by the energy of fission a 15-word buffered interface into the core of a Perkin-Elmer fragments from the spontaneous fission of californium-252. Interdata 8/32 computer. Up to 32,000 TOF data points were 252Cf-PDMS is particularly useful for determining the mass of recorded simultaneously. After data acquisition the data were very large ions, and ions of mass greater than 10,000 daltons analyzed interactively with a Tektronix 4014-1 graphics ter- have been observed (18). We report here data on the compo- minal. The analysis involved mass calibration, background sition of self-assembled Chl a dimers and oligomers, with results subtraction, a peak search, TOF centroid measurement, and that show 252Cf-PDMS to be well suited to the study of self- mass calculation. A mass calibration was calculated for each assembled chlorophyll systems. sample by determining the precise TOF of the H+ and Na+ in the positive ion spectrum and H- and C2H- in the negative ion 252Cf-PDMS spectrum. Using these results, the slope and y intercept of the TOF vs. N/Ki curve were evaluated with a precision of 1 part Instrumentation. A general description of the 252Cf plasma in 35,000. Spectra were recorded with a mass resolution of 350 desorption time-of-flight (TOF) mass spectrometer has been (full width at half maximum). Above M = 700, the '3C satellites given (19). The system was optimized for high-mass ion analysis were unresolved and the measured TOF values of peaks above by reducing the flight path to 45 cm and operating the channel this mass were for the isotopically averaged distribution for a electron multiplier array (CEMA) detectors near saturation for particular ion. The accuracy for measuring averaged masses improved detection efficiency of high-mass ions. The samples was studied by using the chlorophyll monomer ions. A value were irradiated for periods up to 18 hr with a 252Cf source of 893.5 i 0.2 (mean + SD) was measured, which compares giving a flux in the samples of 2500 fission fragments per with a literature value of 893.503 for the isotopically averaged cm2/sec. Positive or negative ions desorbed from the fission mass. For the range above M = 1000, the average accuracy of track were accelerated to 10 keV energy and masses were the mass measurements was found to be 1 part in 5000. In spite measured by the TOF method. of the fact that the mass resolution is less than 1 part in 5000 in "Time-zero" signals were produced by detecting the com- the mass spectrum, the time centroid of the isotopically aver- plementary fission fragments, using an electron conversion foil aged ions can be measured with sufficient accuracy (0.125 psec) and accelerating the electrons into the face of a two-detector to make possible mass measurements accurate to within 1 part CEMA system operating in tandem. in 5000. A diagram of the ion source is shown in Fig. 2. The fission flux Sample Preparation. Chlorophyll samples were dried by into the time-zero detection system was electronically resolved codistillation five times with dry carbon tetrachloride and then from the higher intensity a-particle flux by making use of the degassed for 1 hr under reduced pressure [approximately 10-5 higher electron multiplicity of the fission fragment track in the torr (1.3 mPa)] at room temperature. Solvents were thoroughly electron conversion foil. The CEMA ion detector at the end of dried over molecular sieves and degassed on a vacuum line. the flight tube ("stop" detector) was also operated in a single After drying neither the solvent nor the chlorophyll was directly ion mode so that an ion striking the detector produced a single exposed to air, and all subsequent manipulations were carried transient electronic pulse. Pulses from the time-zero and stop out under dry nitrogen. To prepare oligomers dry Chl a was detectors were amplified and coupled to fast constant-fraction dissolved in n-octane (99%, Mallinckrodt) at a concentration discriminators (ORTEC model 473A), which generated a train of 0.1 M. Films of Chi a (250 Mg/cm2) were prepared on 1- of fast logic time-zero signals and time-delayed stop signals. The gm-thick nickel foils over an area of 1 cm2 by evaporation onto intervals between the time-zero and stop signals were measured the pre-dried nickel foil under a dry nitrogen atmosphere. digitally by using an EG&G time-interval counter (TDC-100) Pheophytin films were prepared in a similar manner except with multiple stop-signal capability. When a particular fission that carbon tetrachloride (Burdick and Jackson, Muskegon, MI) track produced several ions of different mass (M), the TOF of was used as a solvent because pheophytin has only limited sol- each ion was measured with a 4-psec dead time between stop ubility in aliphatic hydrocarbons. Where possible, the elec- signals. TOF measurements were made with 0.125-nsec reso- trospray technique was employed for sample preparation be- lution for mass calibration and 1-nsec resolution for the mass cause it produced more uniform films (20). Solutions suitable spectrum. for the preparation of monomeric Chl a films were prepared in wet benzene. The aggregation state of chlorophyll in thin films was assessed by absorption spectroscopy of thin films deposited on glass slides. RESULTS AND DISCUSSION The 252Cf-PDMS of Chl a films (Fig. 3) shows positive and Electr Grid negative molecular ions (M+ and M-). An intense positive and collector Grid- .i negative ion fragmentation pattern results from successive losses

f'CKAULMAHA of side chains from the intact chlorin ring; the central Mg atom detector :Cf is not lost. A significant fragment ion (M = 615) in the positive assembly Acceleration Electron grid. and negative ion spectra corresponds to the loss of the phytyl conversion ground potential group from the propionic acid side chain at carbon 7. The foil Sample holder formation of both M+ and M- ions of Chl a in the plasma assembly. created by the fission track indicates that rapid redox reactions 10 kV between chlorophyll molecules can occur under these condi- FIG. 2. Diagram of 252Cf-PDMS source region. Fission fragments tions. The extensive fragment ion spectrum suggests that vi- are detected by accelerating secondary electrons into CEMAs after brationally excited Chl a molecular ions are also produced in the electrons are ejected from a foil behind the 252Cf source. Targets are mounted on an eight-position slide. Ions formed at the surface of high yield on a time scale comparable to the fission track life- the sample are accelerated to 10 keV energy and pass through the time (:t;10 psec) (21). This general fragmentation pattern and acceleration grid. the monomer mass spectrum were observed for all Chl a films Downloaded by guest on September 26, 2021 Chemistry: Hunt et al. Proc. Natl. Acad. Sci. USA 77 (1980) 1747 film (Xmaxm 675) prepared from dry carbon tetrachloride A 4 I1 shows an intense dimer ion peak at 1787 daltons (Fig. 4B), .3 which was not present in the mass spectrum for a Chi a film cast 0 from wet benzene. It is known from NMR, infrared, and vapor x phase osmometry measurements that carbon tetrachloride so- lutions of ChG a contain a high concentration of Chi a dimers, (Chi a)2, (8, 22). Thin solid films of Chi a prepared from carbon tetrachloride solution thus appear to retain the dimer self-as- sembly. The dimer exhibits an intense fragmentation pattern in the positive ion spectrum corresponding to the loss of one (M '1542 556 = 1509) and two phytyl groups (M = 1236). Some of the side groups of the chlorin rings are also lost, resulting in a broad 893 distribution of ions centered around M = 1220. The observation of a dimer less both phytyl chains (M = 1236) demonstrates that 400 500 600 700 800 900 even such extensive fragmentation of a dimeric chlorophyll Mass, daltons species does not necessarily disrupt the coordination interaction FIG. 3. 252Cf-PDMS positive (A) and negative (B) ion mass that forms the dimer. spectra of a Chl a film deposited from wet benzene, recorded for In 252Cf-PDMS of other large organic molecules, dimers and 10,000 sec at 2 nsec per channel. Experimental masses of major ions trimers are often observed, usually with intensities orders of appear beside the peaks. magnitude lower than the monomer ion. Fragmentation of these oligomers normally involves dissociation to lower order prepared from different solvents and appear to be independent oligomers and monomer. For the Chl a dimer, the ion yield is of the degree of self-assembly of the chlorophyll in the solution 50% of the monomer yield. Fig. 4C shows the positive ion from which the film was cast. spectrum in the same mass region of Pheo a cast from carbon The presence of oligomer (ChI a). ions in the mass spectrum tetrachloride solution. Here there is no evidence of dimer ion was found to be sensitively dependent on the degree of oh- formation. Thus, the 252Cf-PDMS measurements show that the gomerization of the chlorophyll in the solution from which the central Mg atom must play a dominant role in stabilizing the films were cast, as determined by optical absorption spectra. Chi a dimer structure in solid films. The 7r-7r interaction that Fig. 4A shows the spectrum for a monomer ChI a film deposited is operative in both Pheo a and Chl a does not appear to be from wet benzene. This film has a visible absorption maximum sufficiently strong to account for the stability of the dimer ion. in the red at 662 nm as expected for monomeric chlorophyll. It is interesting to note that the Pheo a spectrum contains (M Only the chlorophyll molecule ion, M+ is seen in this spectrum + H)+ and (M - H)- ions, whereas the Chl a spectrum contains in the mass region 750-3000 daltons. The spectrum of a Chi a both positive (M+) and negative (M-) molecules. The ionization process for Pheo a is proton transfer, whereas for ChI a, elec- (Mt) trons are transferred, prefiguring in a sense the roles of Chl a and Pheo a in the primary events of photosynthesis. It is known from other studies (22) that the longest oligomer chains, (ChI a)., n > 20, are formed in solution in dry aliphatic A hydrocarbons such as n-octane (8). We have now been able to verify the presence of long-chain oligomers in Chl a solutions in n-octane by 252Cf-PDMS. Fig. 5 shows the positive ion spectrum extending to M = 8000 in thin films of ChI a prepared from n-octane, in which Chi a is known from direct molecular 0 893 (MS) weight measurements to be present as large oligomers formed x Z2 by as many as 20 Chi a molecules (22). In addition to the dimer c c ion, the trimer, tetramer, pentamer, and hexamer are observed a 1787 (2M:)

."1 0 B 7Mt 5M 6M+ 6252 + 4

0C O 0 1000 1250 1500 1750 2000

871 (MS) 2Mt 1787.6 _0.7 4500 5400 6300 0 0 1 C x_42679S0.6± 4M1 5 1000 1250 1500 1750 2000 Mass, daltons FIG. 4. 252Cf-PDMS positive ion spectra of Chl a and pheophytin 2000 3000 4000 5000 6000 7000 8000 a. (A) Monomer film deposited from wet benzene recorded for 7200 Mass, daltons sec at 4 nsec per channel; (B) dimer film cast from dry carbon tetra- FIG. 5. 252Cf-PDMS positive ion spectrum of a Chl a film cast chloride recorded for 28,800 sec at 2 nsec per channel; and (C) pheo- from dry n-octane, recorded for 50,000 sec at 8 nsec per channel. The phytin film from dry carbon tetrachloride for 18,000 sec at 2 nsec per self-assembled oligomer ions extend to the heptamer ion. Masses are channel. mean + SD. Downloaded by guest on September 26, 2021 1748 Chemistry: Hunt et al. Proc. Natl. Acad. Sci. USA 77 (1980) with decreasing intensity, and there is some evidence of the 5. Brody, S. S. & Brody, M. (1963) in Photosynthetic Mechanisms heptamer at M = 6252 b 4. The relative intensities of these of Green Plants (Natl. Acad. Sci., Washington, DC), Publ. 1145, oligomer ions do not reflect the concentrations of these species pp. 455-478. in the film. The ion desorption process that occurs in 252Cf- 6. Brody, M. (1968) in The Biology of Euglena, ed. Beutow, D. M. (Academic, New York), Vol. 2, pp. 709-727. PDMS is probably the result of the generation of an electronic 7. Katz, J. J. (1968) Dev. Appl. Spectros. 6, 201-218. shock wave at the surface of the film (23). The considerable 8. Ballschmiter, K., Truesdell, K. & Katz, J. J. (1969) Biochim. translation energy imparted to the molecular ions (m-3 eV) Biophys. Acta 184, 604-613. probably leads to dissociation of some of the larger oligomers 9. Katz, J. J. (1973) in Inorganic Biochemistry, ed. Eichhorn, G. L. into smaller units. No oligomers above the trimer were observed (Elsevier, Amsterdam), Vol. 2, pp. 1022-1066. with Chl a films prepared from carbon tetrachloride solu- 10. Closs, G. L., Katz, J. J., Pennington, F. C., Thomas, M. R. & Strain, tion. H. H. (1963) J. Am. Chem. Soc. 85,3809-3821. The measured molecular masses of the oligomer ions are 11. Scheer, H. & Katz, J. J. (1975) in and Metallopor- shown in Fig. 5. These values are accurate to within 1 dalton phyrins, ed. Smith, K. (Elsevier, Amsterdam), pp. 462-468, i 493-501. except for the weak heptamer ion, for which the accuracy is 12. Katz, J. J., Dougherty, R. C. & Boucher, L. J. (1966) in The 4 daltons. The measured values are in agreement with those , eds. Vernon, L. P. & Seely, G. R. (Academic, New calculated for self-assembled species consisting only of Chl a York), pp. 185-251. molecules with no other ligands incorporated or attached. 13. Shipman, L. L., Janson, T. R., Ray, G. J. & Katz, J. J. (1975) Proc. In conclusion, we have been able to obtain mass spectrometric Natl. Acad. Sci. USA 72,2873-2876. data on self-assembled aggregates of anhydrous Chl a by 14. Kuhn, H. (1979) in Light Induced Charge Separation in Biology 252Gf-PDMS. Our results are entirely consistent with earlier and Chemistry, eds. Gerischer, H. & Katz, J. J. (Verlag Chemie, deductions from solution studies on Chl a dimers and oligomers. Weinheim, West Germany), pp. 151-169. The mass spectrometric data show that, contrary to opinions 15. Katz, J. J., Norris, J. R., Shipman, L. L., Thurnauer, M. C. & in the literature a free of water can Wasielewski, M. R. (1978) Annu. Rev. Biophys. Bioeng. 7, expressed (24-27), (i) Chl 393-434. be readily obtained by simple procedures and (ii) the Chl a 16. Katz, J. J., Oettmeier, W. & Norris, J. R. (1976) Philos. Trans. R. species absorbing maximally at t-680 nm contain no water- Soc. London Ser. B 273, 227-253. i.e., water is not an integral component of the (Chl a). oligomer 17. Shipman, L. L., Cotton, T. M., Norris, J. R. & Katz, J. J. (1976) structure. J. Am. Chem. Soc. 98,8222-8230. 18. Macfarlane, R. D. & Torgerson, D. F. (1976) Science 191, 920-925. The authors thank Mr. R. A. Martin for support in computer appli- 19. Macfarlane, R. D. & Torgerson, D. F. (1976) Int. J. Mass Spec- cations and Mr. S. Williams for general technical assistance. This re- trom. Ion Phys. 21, 81-92. search was supported by the National Science Foundation (CHE- 20. McNeal, C. J., Macfarlane, R. D. & Thurston, E. L. (1979) Anal. 7804863) (R.D.M.); the Robert A. Welch Foundation (J.E.H., R.D.M.); Chem. 51, 2036-2039. The Division of Chemical Sciences, Office of Basic Energy Sciences 21. Seitz, F. & Koehler, J. S. (1956) Solid State Phys. 2,307-371. of the U.S. Department of Energy (.J.K.); and the National Institutes 22. Katz, J. J., Shipman, L. L., Cotton, T. M. & Janson, T. R. (1978) of Health (R.C.D.). in The Porphyrins, ed. Dolphin, D. (Academic, New York), Vol. 5, Part C, pp. 401-458. 23. Macfarlane, R. D. (1979) National Bureau ofStandards Special 1. Rabinowitch, E (1951) Photosynthesis (Interscience, New York), Publication 519, pp. 673-677. Vol. 2, Part 1, pp. 697-701. 24. Fong, F. K. & Koester, V. J. (1975) J. Am. Chem. Soc. 97, 2. Katz, J. J., Closs, G. L., Pennington, F. C., Thomas, M. R. & Strain, 6888-6890. H. H. (1963) J. Am. Chem. Soc. 85, 3801-3809. 25. Fong, F. K. & Koester, V. J. (1976) Biochim. Biophys. Acta 423, 3. Rabinowitch, E. & Govindjee (1969) Photosynthesis (Wiley, New 52-64. York), pp. 196-215. 26. Winograd, N., Shepard, A., Karweik, D. H., Koester, V. J. & Fong, 4. Krasnovskii, A. A. (1969) in Progress in Photosynthesis Research, F. K. (1976) J. Am. Chem. Soc. 98,2370-2371. ed. Metzner, H. (International Union of Biological Sciences, 27. Brace, J. G., Fong, F. K., Karweik, D. H., Koester, V. J., Shepard, Tfibingen, West Germany), Vol. 2, pp. 709-727. A. & Winograd, N. (1978) J. Am. Chem. Soc. 100, 5203-5207. Downloaded by guest on September 26, 2021