MSU LIBRARIES “ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. filfl§§_will be charged if book is returned after the date stamped below. THE DEVELOPMENT AND CHARACTERIZATION OF A FLASH-KINETIC ABSORPTION SPECTHOHETER by Dwight Henry Lillie A THESIS Submitted to lichignn Stnte University in pnrtinl fulfillment of the requirements for the degree of MASTER OF SCIENCE DEPARTIENT OF CHEMISTRY 1984 3 (7‘7“: TEL) ‘93“; ABSTRACT THE DEVELOPMENT AND CHARACTERIZATION OF A FLASH-KINETIC ABSORPTION SEECTROMETER By Dwight Henry Lillie A computer-based modular optical kinetic spectrometer has been deveIOped for the purpose of studying some of the primary reactions in green plant photosynthesis. This spectrometer has the capability of detecting Optical absorption changes on the order of 10-4. and can follow reactions with a time constant of 2Mhz. The modularity of the system allows for differing experimental setup requirements. A series of experiments have been performed to ascertain the spectrometer's current capabilities and limits. These involved, among others. the induced photochemistry from the probe source, electrochromic response seen in chlorOplasts. proton release as measured by an indicator dye, Neutral Red. Results indicate the kinetic spectrometer reproduces the same as in the literature. we now have the capability to probe optically the oxidizing sire of PSII; both at Z the primary donor to P680 and at the Oxygen Evolving Complex. to Carolynn. Curtis, and my parents -11— ACINOILEDGEMENTS I would like to thank Dr. Gerald T. Babcock for the independence and assistance he offered throughout this project. I am grateful for the patience and time Ron Haas. Tom Clarke, and Marty Rabb spent teaching and helping me with the electronics. I am grateful to Dr. Tom Atkinson for the help in programming and de-bugging. Finally. I thank my wife. Carolynn, for her patience and love. -iii- TABLE OF CONTENTS Chapter LIST OF FIGURES ABBREVIATIONS 1. INTRODUCTION 1.0 PREFACE 1.1 PROTDSTN'HIESIS. AN (NEVIEI 1.2 OPTICAL ABSORBANCE THEORY 1.3 EXPERIMENTAL APPROACHES 1.3.1 Optical 1.3.2 Detection 2. INSTRUMENT DESIGN 2.0 SUBSYSTEMS 2.1 OPTICAL 2.2 ELECTRONIC 2.2.1 Detector Amplifiers 2.2.2 Timing Circuitry 2.3 COMPUTER 2.3.1 General 2.3.2 Nicolet 1074 2.3.3 AIM-65 Microcomputer 3. SOFTIARE DESIGN 3.0 PROGRAMMING CONSIDERATIONS 3.1 AIM-65 3.1.1 System Execution 3.1.1.1 Program Description 3.2 LSI-11/02-23 4. EXPERIMENTS AND FUTURE PLANS 4.0 MATERIALS AND METHODS 4.0.1 Chloroplast Preparation 4.0.2 Standard Experimental Conditions 4.1 CONTROL EXPERIMENTS 4.1.1 Steady State Effect of Probe Beam 4.1.2 Flash-induced Difference Absorption Spectrum .. 1v... PAGE 2 Chapter Page 4.1.3 Effect of DCMU on the 518mm Change 77 4.1.4 Neutral Red Absorbance Changes 80 4.2 PLANNED EXPERIMENTS 85 4.2.1 Spectral and Kinetic Identification of Z 85 APPENDIX A 87 LIST'OF REFERENCES 101 -r Figure 1.1 1.2 1.3 1.4 1.5 2.1 LIST OF FIGURES Z-SCHEME in green plants. Display of the linear electron transport pathway as a function of the known or inferred midpoint potentials of the redox couples involved. Patterns of electron and proton transport in relation to the thylakoid membrane. Z-SCHEME. Depiction of known absorbance changes and the time constants associated with the components of the photosynthetic electron transport chain. OXYGEN FLASH YIELD. The oxygen flash yield in isolated chloroplasts after a long dark time as a function of the number of flashes. Xok et a1. (43) were able to fit a model assuming a. a fraction of centers not undergoing S _> S n+ transition. and B, a fraction undergoing double transitions. m) Sn 2. Computer modeling predicts dark-adapted cfiloroplast OECs to be 75$ 81 and 25% So (68). Influence of an electric field on the transition energy from S . ground state to S . an excited state. (A) Indicates indicates the change of the transition enery as the field is turned on . (B) The absorption band shift (above) is depicted and the net result is the derivative shape (below). and is actually seen (FIGURE 4.4). OPTICAL LAYOUT of the FAS. Only relative positions are shown. distances are not to scale. -Vi- Page 10 13 18 31 Figure 2.2 2.3 2.4 2.5 2.6 ELECTRONIC LATOUT of the FAS. Analogue data pathways are given in solid lines. and the digital data lines are dashed. Only the electrical pathways are shown here. Note the electrical isolation of the Laser and it's trigger circuit all other electronic components of the FAS run to a common ground (not shown). DC LIGHT MODIFICATION. This layout modifies the 28V AC current in to 25V DC C1 8 2400 “F; C2.C3.C4 8 10 uF; R1 = 120 ohm; R2 - 5 [ohm . variable; Dl.D2 = IN4002. protection diodes; LM338 5A.25V power regulator (National Semiconductor Corp.; Santa Clara.CA). 600nm CUTDFF FILTER SPECTRUM. Two such filters are used to guard the PMT from laser scattering. PMT CASCADE AMPLIFIER CIRCUIT. The first stage is connected to the base of the PMT by a 6cm long . heavily shielded, conductor. This is to reduce capacitance effects and pickup of Radio Frequency. The second and third stages are then housed in a separate box. Final output runs to a sample-and-hold amplifier. Adjustable gain on the thirds stage gives overall gains between 0.1 and 10,000. For a gain of l. the adjustable time constant ranges between 3.8Mhz and 1H1. SAMPLE-and-HOLD AMPLIFIER (S/H). The other two amplifiers after the SHArS are LF351s (not labelled). The input line is divided with one lead going to the SHAPS and into the first LF351's inverting input. and the second lead ties up in a voltage summing configuration going into the scond LF351's inverting input. 'hen the SHArS is in the SAMPLE mode it passes the input signal completely. it's inverted and summed with the positive signal for an output signal of 0.0V. Ihen the SHArS is triggered into the HOLD mode it stores the voltage on a capacitor and holds that voltage on its ouput line. it is inverted. and subtracted from the -vii- Page 33 37 39 42 45 Figure 2.7 2.8 4.1 4.2 present signal resulting in an output signal centered around 0.0V D.C. TIMING CIRCUITT. The AIMr65 is required to bring four address lines. A0 - A3, and a timing line (labelled TTL). in to run the circuitry. The 74C154 is a 4 to 16 decoder and provides 16 discrete output lines (only 12 are currently utilized). The 7400 (NAND gate) and 7404 (nor INVERTER) are used to shape each transitions on the TTL line into pulses (detailed description in text and logic involved in FIGURE 2.8). A monostable vibrator (74121) trims the pulse duration to 2.0 :s. and the output is fed into input of each of 7402 (NAND gate). The NAND gate selected will be the one which has its input line from the decoder L0 when the other input. from the monostable. goes L0 it will produce an output HI pulse of the same duration as the monostable's. TIMING SCHEMATIC. The logic involved in transforming each transition (LO -) HI or HI -> L0) on the input TTL line. A. into a final output line. F. with one pulse per transition. The output of each gate is labelled and is matched with a time trace below. OXYGEN EVOLUTION. This depicts typical oxygen evolving rates seen with class II chlorOplasts prepared as described. These rates were obtained with freshly prepared chlorOplasts. Rates obtained with freeze-thawed chloroplasts (not shown) are similar. (A) Coupled oxygen evolving rate. (B) Uhcoupled oxygen evolving rate. A typical sequence of timing events (not to scale). On the tap line. A. each ti indicates the beginning of a timing cycle. The cycle repeats itself at every t . the smaller line after the t marker indicates t at the active event timing is comlpIeted. and the AIM-65 enters an idle 100p until the beginning of the next cycle. Line B is an expansion of one cycle, and line C is an expansion of the active event timing portion. Example times are given below -viii- Page 49 51 66 7O Figure 4.3 4.4 4.5 4.6 (note: all times are given with respect to the previous event) : time dgration event tia 50ms photoshutter cpen tib 10ms S/H - 'hold' tic 0-lms A/D - begin converting tid 1-20ms NICOLET - begin storage tie 10ms S/H - 'sample' tlf 1ms photoshutter close tlg l-lSs enter idle loop PROBE LIGHT EFFECT. The effect of contiuous illumination of the sample by the probe for: (A) 40.92 seconds. and (B) 1.0 second. FLASH INDUCED DIFFERENCE ABSORPTION SPECTRUM. Center line represents zero absorption changes. Each + marks one point. EFFECT OF DCMU ON 518 CHANGE. Baselines of the two traces shown were overlapped to emphasize the difference in absorption changes. Trace A is a normal 518 change. and trace R is the response with 166 u! DCMU. The negative spike(s) are at time zero. and is the PMT's response to the laser excitation. Time constant was less than 5 ' 10 sec. [DCMUJ-166 “M. [MVJ-49 “M. and equilibration for 60 seconds was allowed. NEUTRAL RED ABSORBANCE CHANGES. As discussed in the text. this is the result of subtracting the absorption changes at 529nm with 40 u! N.R. with those obtained without N.R. in solution. Both original data sets contained sharp negative peaks at time zero. subtraction took most of it out. Each data set also contained laser-induced scatter. Subtraction of one from another automatically eliminates the scatter. —. ix- Page 72 75 78 81 BSA BCD Chl DEC FSK NADPH OEC PSI PSII «1,10,, ABBREVIATIONS Adenosine Triphosphate Bovine Serum Albumin Binary Coded Decimal Chlorophyll Digital Equipment Corporation POP-11, LSI-ll. PDP. or LSI are often used in referring to a DEC computers Dibrmo-3-methyl-6-isopropyl-l.4 benzoquineone 3-(3.4 Dichlorophenyl)-1.1 dimethylurea Ethylenediamine tetraacetic acid Electron-Nuclear Double Resonance Electron Spin Resonance; often called EPR for Electron Paramagnetic Resonance Frequency Shift Xeyed N-z-Rydroxyethyl piperarine-N“-2-ethanosulfonic acid Microprosser Unit; often called CPU for Central Processing Unit Methyl Viologen Nicotinamide adenonine dinucleotide phosphate Neutral Red Oxygen Evolving Complex Plastoquinone Photosystem I Photosystem II quinone species ivolved in electron transport chain between PSII and PSII RSXIIM RT‘11 SHN S/H SIN TRIS Read Only Memory Random Access Memory Real time multi-tasking. multi-user operating system supported by Digital Equipment Corporation. Designed to operate on their PDP-ll mini-computers. Real time single user operating system supported by Digital Equipment Corporation. Standard Deviation A solution of sucrose,helpes. and salt Sample and Hold Amplifier Signal-to-noise ratio Tris(hydroxymethyl)aminomethane Primary donor to P680. also thought to be origin of Signals IIvf and IIf CHAPTER 1 INTRODUCTION 1.0 PREFACE Researchers working on solar energy conversion systems would do well to pay heed to what green plants do so well. Splitting water. driving electrons ”uphill" over one volt. and generally getting usable chemical energy from sunlight is not a trivial feat! Plants require the cooperation of two photons of differing energies (1.85 eV and 1.77 eV) and two serially linked light-to-energy conversion systems to produce useful energy. In the process, through their highly structured conversion system. natural photosynthetic systems can boast of the highest sustained electron transfer rates ever measured. and theoretical studies by Ross et al. (17) indicate that the maximum efficiency is 70%. As 60 to 70% of the energy of the incident photons is used for driving a charge separation across a membrane; total yield for natural photosynthetic systems approach 90$ of the theoretical upper limit! 1.1 PHOTOSTNTHESIS, AN OVERVIEW The general net reaction of photosynthesis can be written as: C02 + HQA --) (CHZO) + 02. (1.1) where H2A is an oxidizable substance such as H2.H20,H28. simple alcohols. or fatty acids (in the latter two cases HzA is also often the source of carbon for the reaction); CH20 is stored as an energy supply by the biological system. The specific form of Equation 1.1 for green plants is: In higher plants these reactions occur in subcellular organelles known as chlorOplasts. vhich are 10-50 micrometers (um) in diameter. and consist of an inner and an outer membrane. The inner membrane is continuous and typically forms flattened vesicles (0.5 pm in diameter) which stack upon each other in configurations known as grana. and the outer membrane is typically sheared during preparation of class II chloroplasts. The grana are interconnected by unstacked membranes known as the stroma lamella. "Thylakoid" is the term for the individual vesicles in the grana; conventionally. the terms: class II chloroplasts. thylakoids. and chloroplasts are used interchangeably. Equation (1.2) can be misleading in its simplicity and lack of mechanistic information as there are two distinct sets of chemical reactions which occur in photosynthesis. Light reactions. catalyzed by light. perform the actual conversion to chemical energy in the form of nicotinamide adenine dinucleotide phosphate (NADPH) and adenosine triphosphate (ATP). The molecular structures responsible for the light reactions are located in and span the width of the thylakoid membrane. Dark reactions. otherwise known as the Calvin cycle (12). use the products of the light reactions in the reduction of C02 to carbohydrate. and are located in the stroma lamella. For a general introduction see reference (16). The current nomenclature and representation of the structures involved in the light reactions are depicted in Figure 1.1. categorized by the midpoint potentials of the redox couples. Known as the Z-scheme. based on a model prepsed in 1960 by Hill and Bendall (13). it has been elaborated since by Hind (14) and Avron (15). and generally accepted as an accurate model for electron flow from 2E20/05+4HI to NADPINADPH. Two photosystems. photosystem I (PSI) and photosystom II (PSII). have been shown to be linked in series (27) so that under physiological conditions electron transfer is linear. proceeding from the inner thylakoid membrane to the exterior. Figure 1.2 depicts the general orientation of the Z-scheme components in the membrane. P680 AND P700. the primary donors or reaction centers (RC) of PSII and PSI respectively. undergo excitation and photooxidation upon exciton transfer mechanisms into them by the chlorophyll light-harvesting systems (for a review. see 28.29). P680 and P700 undergo absorbance bleachings around 680 um and 700 nm. respectively. Until recently both P680 and P700 were thought to be chlorOphyll dimers (30.31) because of similarities in electron spin resonance (ESR) and electron-nuclear double resonance (ENDOR) spectra of the .13‘1139 reaction centers and dimer chlorophyll-water adducts. Reevaluation of the data. new evidence (32.33). and consideration given FIGURE 1.1 Z-SCHEME in green plants. Display of the linear electron transport pathway as a function of the known or inferred midpoint potentials of the redox couples involved. o _ md 0.0 m o- o._ .. q d. J u d 80.. v 2:: : \uL» / \ Z. w ( _Cm_ m0e\o Guru No ++rv so. \ o... ..\ c... “6.. . OOhm vx ....\s. .0 .\ a n. amouuawom .\ +842 FIGURE 1.2 Patterns of electron and proton transport in relation to the thylakoid membrane. .Im a ONT—C + a2: '7 5:. . i: + +QO 0680+PheoQAQB-. (1 .6) GB will accept two electrons before exchanging with a ”pool" of 7-10 other PCs in the membrane. After the PO pool a series of other carriers finally reduce P700+' Subsequent oxidation of QA- is between 200-400 and 600-800 us (Recall that QB is a two electron acceptor. Upon accepting the second electron. it leaves its site and another P0 must diffuse in before additional electron transfer can occur.). P680+ is reduced by Z (41) (also referred to as D(39)) which is in turn reduced by the oxygen evolving complex (OEC) and represented as Sn in reactions (1.7) and (1.8): + + snzP6SO Phoo “-“> Snz P‘soPhQO. (1.7) + Sh; P‘sophoo -') Sn+IZP680PhOO. (1.8) Reduction times of P680 are dependent upon sample preparation. handling. and the type of experiment performed. Single flash experiments upon dark-adapted chloroplasts by van Rest and Mathis (18) yield the fast reduction times of 30 ns. Steady state flash experiments (21) report a much slower time for reduction of 600 ns. Figure 1.3 provides an overview of the Z-scheme as a function of the time scales involved and the Optically detectable components. Oxidation of 2 occurs during the reduction of P680+ and its reduction proceeding at a rate determined by the redox state of the OEC. The OEC is an enzyme system probably containing four active or participating manganese atoms thought to be responsible for the oxidation of water. Under flashing light Joliet et al. (42) observed 10 FIGURE 1.3 Z-SCHEME. Depiction of known absorbance changes and the time constants associated with the components of the photosynthetic electron transport chain. 11 . m3 V can cue cum 355222.; 23.3 3o 2% Se . 3m .55.: N y 0.“ + +1 E C l—OOn— ‘ ( : Twp...) Nlomcnloognrli On 3 of as: 2.98.6.» . T53.3« 268-2 32$ c2525. 2.509. 2. anoou 12 that the 02 yield per flash varied with a period of four (Figure 1.4). Rok et al. (43) preposed a cyclic model for the OEC: 30-!“ 3,4!» 32-94 334'!» s. (1.9) 4H+ + 02 2H20 The OEC cycles through four oxidizing equivalents to oxidize two water molecules. releasing four protons to the chloroplast interior for every molecule of oxygen produced. No solid evidence is available concerning which S state transition releases protons. nor is there anything more than speculation available concerning the molecular structure of the "oxygen precursors." and their possible orientation relative to the manganese. Extensive use of neutral red. a membrane permeant pH indicator dye. has been extensivly used by Hong and Junge (57). Theg and Junge (58). Xhanna et al. (59). and Velthuys (60) to characterize proton release. Junge performed a rigorous set of control studies (61) showing that neutral red is a rapid indicator for proton release to the interior of the chloroplasts. However. wide debate continues in the literature between Fowler (46). Junge (44). Renger (45). and others concerning proton release patterns. sites in the electron transport chain where protons might be released. and involvement of manganese in water oxidation. Johnson et a1. (62) have recently shown that the chloroplast 13 FIGURE 1.4 OXYGEN FLASH YIELD. The oxygen flash yield in isolated chloroplasts after a long dark time as a function of the number of flashes. Xok et al. (43) were able to fit a model assuming a. a fraction of centers not undergoing Sn --> Sn+1 transition. and B. a fraction undergoing double transitions. Sn --) Sn+2. Computer modeling predicts dark-adapted chlorOplast OECs to be 75$ S1 and 25% So (68). 14 ON E0232 :3 O p 0.0 0d 0.? 15 membrane might be a proton "sink." being able to release at least 1 H+ per 10 chlorophyll. It is conceivable that a parallel phenomena exists in hemoglobin. nmmely the Bohr effect (47). in which the binding and release of oxygen changes the p! of ionizable groups in the protein causing them to release or take up protons accordingly. Z itself has been directly detected only recently. ESR transients. signals IIvf and IIf in oxygen evolving and oxygen inhibited chlorOplasts respectively. detected and characterized by Blankenship. Rabcock. and Sauer (48.49.50) have been assigned to be Z'+ since 1975. The kinetics of 2 should be inferable from the necessary correlation with the rapid absorption changes of P680+. The ini1ysii of 211‘,f ind 11f reveals such a correlation. Babcock et a1. (41)proposed for the various oxidation states of the OEC the following scheme: z-+ + so --> z + 81 t1,2 S 100 us; (1.10) 2.. + 31 -—> z + 82 t1,2 s 100 pi'1’; (1.11) Z-+ + 82 --) Z + S3 t1/2 C 400 us; (1.12) Z-+ + S3 --> Z + S4 t1/2 ' 1 ms; (1.13) Renger and Voelker (52) have recently been able to correlate proton P630+ release transients with the extent of oxidation. Since then. Renger and Ieiss (51) have observed an cptieal transient with a ‘1)Delayed light and electric field measurements on chloroplasts by Duttner and Rabeock (unpublished) appear to indicate a t1/2 of 70 "a 16 periodicity of four at 320 um under a flashing light; the phenomena appears to be related to Z. Dekker et al (76) have apparently obtained the most convincing optical evidence to date on Z°+. they report bands in the UV at 260 nm. 300 mm. and between 390 to 450nm in Tris-washed PSII preparations. Initial photochemistry. as described reactions (1.5) and (1.6) and Figure 1.3. sets up a charge separation across the chloroplast membrane. Concurrently. large positive absorption changes have been detected between 400 - 565 nm (2). and attributed by Junge and Mitt (54) to the charge separation created by the initial photochemical reactions. The charge separation causes an electric field on the order of 107 V'mrl and can induce absorption band shifts (electrochromism) on bulk membrane pigments such as chlorophylls and carotenoids. In general. a pigment molecule absorbing light goes from a ground to an excited state. Application of an electric field will change the transition energy between the two states if they differ in their dipole moments or polarizabilities. The frequency of the shift (for a review of theory. Liptay (55)) is related to the electric field : Av - -1/h(u' - 31°); - 1/2h(i‘ - i°)g2 (1.14) where h is Planck's constant; p. and no are dipole moment vectors of the excited and ground states respectively:.§ is the electric field strength vector; and a. and a0 the polarizabilities of the excited and ground states respectively. It has been observed that the change in absorbance is linear with respect to the membrane potential (see Reiuwald et al. (70) and Amesz and De Grooth (71)). Carotenoids. 17 however. should display no such behaviour as they do not possess a dipole moment. Schmidt (72) proposed the presence of a permanent local electric field in the membrane causing an induced dipole in the carotenoids. This induced dipole would be responsible for the linear dependence Of the absorbance shifts in carotenoids due to the externally applied fields. Figure 1.5a illustrates the shift in energy levels and Figure 1.5b. the shift of an absorption band induced by an electric field. The 518 change‘l) has a rise time of less than 20ns after a short laser flash and has a decay time Of 10ms - 1s depending upon membrane permeability to ions and the pH gradient across the membrane. Contributing to the total electric field are two photoreaction charge separations from PSII and PSI. Selectively blocking PSI or PSII from Operating decreases the maximum absorbance change by half (56). The electron transfer chain can be altered selectively by inhibitors. The reaction in (1.5) can be blocked by 3-(3.4 dichlorOpheny1)-1.1-dimethylurea (DCMU). 2.5-Dibromo-3-mmthyl-6isoprOpyl-l.4 benzoquinone (DRMIR) blocks Equation (1.6) and washing with Tris(hydroxymethyl)aminomethane (TRIS) during chloroplast preparation blocks Reaction (1.8). Addition of electron donors such as benzidine or ferrocyanide. electron acceptors such as methyl violOgen (MV) or ferricyanide. and a blocker for PSI or PSII allows selective probing Of .1)This refers loosely to a large positive absorption band observed from 515 to 540 nm; in the literature it's referred to as the '515'. the '518'. or '522' change depending. Of course. on the wavelength it's measured at. 18 FIGURE 1.5 Influence of an electric field on the transition energy from So. ground state to 81. an excited state. (A) Indicates indicates the change Of the transition enery as the field is "turned on”. (B) The absorption band shift (above) is depicted: and the net result is the derivative shape (below). and is actually seen (FIGURE 4.4). ENERGY-* 8, .._._.:.--...-T.-- T IwE W“ s _,.o" ~. ' o -—----- ----—— FlELD—> FREQUENCY 19 20 one photosystem or the other. Under altered conditions (the requirementsfor very fast electron transfer may no longer hold true and back or cyclic electron transfer reactions can occur; see for example Hong et al. (53). 1.2 OPTICAL ABSORBANCE THEORY For induced transitions in a molecular system. the incident photon must be of energy equal to the difference in energy between two states. The probability Of the absorption of a photon in unit time Of energy equal to: AB<1-lkn> ' El1n> ’ EIkn)’ ‘1-15’ where Ilm) is a lower vibronic state and (km) a higher one. is: P“<1-lkn) ‘ B<1nlkn) ‘ P‘“<1-lkn>" ‘1-1" where B is the Einstein transition probability coefficient for induced absorption. and P(”). In a similar fashion. the probability Of induced emission can be expressed as : P19 - B . D(U(hl1.>)e (1.17) where B is the Einstein transition probability coefficient for induced emission and is also equal to the coefficient for induced absorption. One other transition. the probability of spontaneous emission. is expressed as : 21 P“(tau-o ' A ‘ P")' (1-13’ where A is the Einstein transition probability coefficient for spontaneous emission. Under steady state or equilibrium conditions the probability of absorption equals the probability of emission Pia . Pie + Pse. (1.19) For populations Of the states llm) and (km) of NIlm) and NIkn)‘ respectively. the total probability for absorption and emission Of radiation of “<1mlkn) is : Ptot = PiO‘N'kn> + Pse‘len) - Piu‘Nlll) (1.201) A‘P‘“’ ’ ("Ikn>'"l1n>’ (1.20b) As an example. consider a system in thermal equilibrium at 298K and a photon at wavelength 600 nm (AE - 2.07eV). The ratio of pOpulations is given by the Boltzman distribution law : "lkn>’"I1-> - o’AE’IT - 9.45 0 10'3“. (1.21) where k is the Boltzman constant and T - 298R. Therefore. the result Of an incident photon in resonance with an energy state of a system is the absorption of radiation and transition to a higher (excited) state. The Einstein coefficients can be determined empirically or calculated from first principles. 22 For flash absorption measurements where essentially two perturbations of the system Occur. the situation is little changed but conceptually harder to visualize. Recall the probability Of a transition is prOportional to the radiation density (and this in turn is prOportional to the energy flux or actual input intensity); if a ”probe" beam's intensity is kept low the ground state (8°)1)" is preferentially populated and competing processes are minimal. Any transitions would only be between ground and the first excited singlet state (8"). An intense exciting flash - as from a laser - would cause a trgn,1tion from S ___> Sn (ground and the nth excited singlet 0 respectively). ‘he probe would continue to cause the co ___> Sn transitions. and possibly Sn --> Sn+1 transitions if the resonance frequency were still the same. Assuming only the So --) 81 transition is active. an increase in the transmitted intensity Of the probe would be detected as the concentration Of the molecular state responsible for the So --) 81 transition would have been depleted. Beer's Law provides a relation between tranomitted light and the concentration of the absorbing species. The differential form : -AI . .(x) . I . C . A1 (1e22) where A1 is the attenuation of light intensity (for this case. it is the probe) per A1 length. a(e) is a proportionality constant related to the Einstein extinction coefficients. I is the initial radiant energy. and C .1)DO not confuse the use of S in referring to molecular states with the 8 states of the OEC discussed earlier. 23 the concentration Of absorbing species. Integration yields : I - I°o"‘“’ 0 c e 1. (1.23) or I - 1°1o" t c 0 1. (1.24) where s is the decadic extinction coefficient. Rearranging yields A - -log(I/I°) - s e c o 1. (1.25) a linear relationship between absorbance and concentration. 1.3 EXPERIMENTAL APPROACHES Norrish and Porter (1) are usually credited with having first successfully Observed fast photolytic transient products Of small gas molecules. Since that time flash experimental protocol has evolved considerably. Electronic detection of events lasting no more than 5 to 10 nanoseconds (us) is practical: however. for sub-nanosecond time resolution researchers are forced to use streak cameras. The trade-Off comes from a decreased sensitivity at the higher speeds as shown later. The basic "classical” setup differs little from conventional UV-VIS spectrometers in terms Of the Optical path and the positions of the components relative to one another. Muny researchers. though. will alter the position Of one or more of the components to achieve Optimal response for the particular application. Junge (2) and Mathis (3) discuss in general the requirements and considerations that go into the design of fast Optical detection equipment. Basically. the designs can "24 be divided into two sub-systems as follows: Optics and detection. As our concern is principally with nano- to micro-second resolution. the discussion addresses this application with digressions where apprOpriate. 1.3.1 Optical - The probe beam. monochromator or interference filter. focusing Optic(s). and detector are usually' mounted on a heavy Optical rail. Exciting source and focusing Optic(s) are placed at right angles to the sample. The heavy Optical rail will minimize effects of any high frequency vibrations and insure stability Of the focus. Distances between sample cell and detector are very critical. ChlorOplasts are highly scattering: incident light will be scattered into the solid angle around the collimated probe beam. As the detector is moved further from the sample it receives less and less of the scattered light and Of the signal of interest. Fluorescence also emits light into the solid angle and becomes a significant portion of the signal detected if the detector is tOO close. thus resulting in an apparent decrease in absorbance. Modulation Of the probe beam and demodulation of the output signal can be done if the interference is too great as fluorescence emission will not contain the modulated component and will be extracted from the output signal when demodulation is performed. Typical modulation techniques are on the order of 100 [Hz while fluorescence lifetimes in chloroplast systems are on the order of 100 MHz. Photons from the exciting source itself. typically a laser. or Team 25 flashlamps. are to be excluded as the detector will respond to all incident photons. Some designs have used a second monochromator with a wide Open bandpass before the detector tO eliminate stray light; however. in highly scattering samples. or when the signal is small. this is usually unfeasible. Even filters between sample cell and detector designed to cut off radiation at the exciting wave length will have an attenuation of only two or three absorbance units (see Figure 2.5). Correspondingly. the transmitted intensity could still be orders Of magnitude more intense than the probe beam. If the detection Of this exciting light were only Of the duration of the flash. which has to be very short in regards to the event under study. then the experiment may be valid. EOwever. if the detector or amplification circuits are overloaded an apparent absorbance decrease can be detected. valid information may still exist within the relaxation curve. but its resolution can be difficult and questionable. 1.3.2 Detection - Primary factors in the design Of electronic detection circuitry are the signal-to-noise (SIN) level. frequency response of the discrete components as well as the system as a whole. gain. stability. and the type Of photodetector used. Van Best and Mathis (4.5) utilize a silicon photodiode (UDT PIN 10. United Detector Technology; Culver City. CA) and are able to detect at least AA - 5.0 9 10'"4 in photosynthetic systems between 515 and 820 nm with an optimal response between 200Hz — 1MHz. Other photosynthetic researchers. Junge (6) and Crofts (7) for example. utilize 26 photomultiplier tubes (PMT). Ingle and Crouch (8) have noted that. in considering the type Of detector alone. except for applications where the probe is of very high intensity. the SIN level Obtained with a photomultiplier tube is greater than with a photodiode. Alternatively. Bernstein. Rothberg. and Peters (9) have described a photoacoustic detection technique applied to low concentration and absorbing transients of benzophenone on a picosecond time scale. Though not applied to photosynthetic systems their results indicate high sensitivity as well as high time-resolution. Other fast nanosecond and sub-nanosecond instruments use a flash probe beam in which the probe beam is pulsed for a very short time at discrete intervals after the exciting flash have been described. Though SIN enhancement is gained over more conventional continuous probe beam instruments. the time required to perform the experiment is much longer than with conventional setups as each datum has to be collected individually. FOr photosynthetic measurements. to insure integrity of the biological system. this requires all experiments to be performed at 0°C or with repeatedly changed sample. In most cases. the maximum frequency response of the system is determined by the amplification circuitry and not the detector. The Gain-Bandwidth Product (GDP) is a constant and determined by the design Of an amplifier. and it essentially gives a measure of the highest frequency response for a particular gain. As an example. the BB3551J (Burr-Brown; Tucson. AR) has a GDP of 50MHz; a gain of 50 reduces the frequency response to 1MHz. To achieve high gain and maximal frequency response several amplifiers are Often set up in series. Known as 27 cascade amplification. the series possesses a maximum frequency response limited by the smallest bandwidth of the individual amplifiers: 2 1In 02 I (2 -1)]'/2 ‘ “I (1.26) where «2 is the high end frequency response. n is the total number of amplifiers. and 01 is the smallest maximum frequency response of n amplifiers. NOise is the major limiting factor in the detection of small fast transients. Every component in an electronic circuit injects a finite quantity of noise into the signal being processed. Minimization of noise without severe signal distortion must Occur in the photodetector. Amplifiers Operating on a signal containing noise will process both equally. A.PMT's response to light is prOportional to the intensity of the light striking the surface. and can be expressed as: i a o e O e At, (1.27) where i is the raw output signal (current); 0 the photon flux upon the photocathode surface; 0 the quantum efficiency Of the PMT; and Ar the time interval. Photons striking the surface Of the PMT are randomly distributed in time. and the PMT's output reflects this. the standard deviation being: 6 - mp)“2 - maul”. (1.28) where c is the standard deviation; and Nb is the number of photons striking the PMT face in the time interval At. NOise from the PMT. in this context. can be defined as the root-mean-square (rms) or standard 28 deviation of the signal. 0. Thus Equation (1.29). sm - i/c - («ml/2. (1.29) provides an expression for evaluation of the SIN. Improvements Of SIN can be achieved by increasing the photon flux across the PMT. increasing the time interval. or using a PMT with a photocathode of a higher quantum efficiency. Equation (1.29) demonstrates why detection of fast transients is difficult. due to short time intervals the SIN is necessarily limited. A factor not considered in this discussion is the Operating temperature Of the system. Electrons are released from the photocathode surface after work has been done to remove them. PMTs maintain very high potential drOps of 500-1500V in order to lower the apparent work function; therefore. a significant fraction of the electrons available at the surface will possess enough energy to leave without benefit of a photon. This is known as "dark current." and can be a significant fraction Of the maximum allowable current of the PMT. Additionally. the noise in the resulting signals is much higher. Mater cooling of the housing of the PMT can reduce dark current noise levels an order of magnitude Of more. As a result Of Equation (1.29) many researchers resort to "signal averaging" to enhance their signal to noise. and the gain to be expected can be expressed as in Equation (1.30): 29 (smog - 11‘1”) e (sum. (1.30) where M is the number of repetitions; (SIN)1 the signal to noise ratio for one flash; and (SIN)! the signal to noise ratio for M flashes. It quickly becomes Obvious that maximal gain Of signal averaging is within the first few iterations. Extremely small signals have to be elucidated in some other manner than signal averaging. In photosynthetic systems experimental times at room temperature cannot exceed 10 to 15 minutes as as irreversible damage to the electron transport chain will then have occurred. CHAPTER 2 INSTRUMENT DESIGN 2.0 SUBSYSTEMS The Flash Absorption Spectrometer (FAS) has three principal systems or levels involved in its design : Optical. electronic. and computer. The following sections will deal with each in a sequential manner; however. considerable overlap exists between the systems. Layouts Of the systems are as in Figures 2.1 and 2.2. 2.1 OPTICAL The Optical subsystem is as in Figure 2.1. All components (except for the PMT) are mounted on an Optical rail bolted to a heavy length of colorlithe (lab bench material). secured to a lab bench top. This insures that high frequency vibrations will not cause mechanical instability in the Optical alignment. A.H-20 monochromator and a 100 watt tungsten light source (both from Instruments SA (ISA); Metuchen. NJ) are used to produce the collimated probe beam. ISA provides an alternating current (AC) variable power. 25 volt. 5 amp. supply for the tungsten light source. but it is completely unacceptable for high precision absorption measurements. Medification to produce direct current (DC) light was effected as in Figure 2.3 with the help Of M. Rabb (MSU Chemistry Department's electronic design specialist) in 30 31 FIGURE 2.1 OPTICAL LAYOUT Of the FAS. Only relative positions are shown. distances are not to scale. 32 22.2541 mmm<4 $5.... mozmmmnmmpz. fl awe-51m mmnm 1.2m m2“: 43.24592 meme-525.05% ozanoo... n -omeummm mzmm i: mafia... Eu... Eocene EEoow 31 FIGURE 2.1 OPTICAL LATOUT of the FAS. Only relative positions are shown. distances are not to scale. 32 29.254»... mwm.x mmoéw>< ._90$ the dye concentration is 5.0 9 10's!_ in methanol. and the minimum time between any two flashes is ten seconds. The high a dye concentration leads to a loss of output energy as rapid energy redistribution takes place among the dye molecules. and too low of a dye concentration leads to higher photolytic rates of decomposition of the dye and loss of power output. The long delay between flashes is necessary as the dye solution must flow down a 6" tube inside of the lamp. Flow rates are kept slow to minimize turbulence which would create inconsistent lasing conditions. Repeatedly lasing the dye without allowing for the tube to be purged and the dye to relax tO the 36 ground state also leads to high decomposition rates and inconsistent lasing action. An initial experiment on green plant photosynthetic systems (for experimental details. see Chapter FOur) tested the saturation of photosynthesis. determined by the electrochromic effect at 518nm. as a function of the laser power output. Variation of the laser output by 50% still continued to fully saturate photosynthesis. The 600nm short-pass filter. the spectrum of which is shown in Figure 2.4. attenuates most of the exciting light at wavelengths higher than 600nm. Additionally. a 640nm interference filter is used in the path of the laser to eliminate white light components. Flashlamp-pumped dye lasers exhibit beam divergence due to a short cavity. thus necessitating two fused-silica focusing Optics to focus the beam upon the sample cuvette. The PMT is an EMT 9558OB (EMI Gencom Inc.. venetian blind configuration. 8-20 response; Plainview. NY) maintained at room temperature at a typical Operating voltage between 530 - 800 V. The RCA 1P28 (RCA. side-on configuration. 8-5 response; Lancaster. PA) was also tried. Per the region Of interest. 300 - 560nm. the 1P28 has a better ability to differentiate between the wavelength of choice and the exciting wavelength in terms of its response curve. At room temperature the 1P28 has less dark current than the 955GB; however. it was unable to detect the fast photosynthetic transients of interest. The initial configuration tried. separated the PMT from the sample cell by 200 cm. A.quartz light pipe was positioned to receive all of the light from the sample and led to the face Of the PMT where a silica lens was used to focus the light emitted from the light pipe onto the PMT. This set-up was used to insure that detection of exciting light 37 FIGURE 2.3 DC LIGHT MODIFICATION. This layout modifies the 28V AC current in to 25V DC Cl - 2400'uF; C2.C3.C4 - 10 uF; R1 - 120 Ohm; R2 - 5 Kohm. R3 - variable; D1.D2 - IN4002. protection diodes; LM338 5A.25V power regulator (National Semiconductor Corp.; Santa Clara.CA). 38 39 FIGURE 2.4 600nm CUTOFF FILTER SPECTRUM. Two such filters are used to guard the PMT from laser scattering. 40 2.4 |.8 |.2 Absorbonce 0.6 0.0 1 I 1 ' l 400 500 600 700 Wavelength (nm) 41 would be minimized; however. detection of the photosynthetic transients was also minimized. At present. 4.5 cm separates the face of the PMT from the sample cell. A new Optical set-up is currently being built which will close the distance between PMT and sample cell to 2.5 cm. Requiring the space around the sample cell and PMT to be completely light tight (except for the exciting and probe sources). easy access to the sample cell. flexibility in choice Of focusing Optics. and easily added or removed blocking filters imposes severe design limitations. especially in terms of compactness. 2 .2 EECI'RWIC 2.2.1 Detector Amplifiers - The three stage amplifier circuit shown in Figure 2.5 is the final result Of much modification and testing. Stages 2 and 3 use the BB3551 amplifier. Its GBP is 5OMHz. Gain on the second stage is 5 thus giving a frequency response Of 10MHz. The third stage is variable gain between 0.1 and 10; experiments usually use gains between 1 and 10. The first stage is a National Semiconductor LF351 (National Semiconductor Corp..GBP of 4MHz; Santa Clara. CA) set in a voltage follower configuration thus leaving the GBP equal to 4MHz. The typical frequency response of the system as a whole by Equation (1.26) is 2MHz. A faster amplifier in the first stage is preferred; however. tests I have performed (data not shown) indicate that the BB3551. if saturated for a short time (nanoseconds). displays a recovery time requiring several 42 FIGURE 2.5 PMT CASCADE AMPLIFIER CIRCUIT. The first stage is connected to the base of the PMT by a 6cm long . heavily shielded. conductor. This is to reduce capacitance effects and pickup of Radio Frequency. The second and third stages are then housed in a separate box. Final output runs to a sample-and-hold amplifier. Adjustable gain on the thirds stage gives overall gains between 0.1 and 10.000. For a gain Of 1. the adjustable time constant ranges between 3.8Mhz and 1Hz. 43 £20m 53-3mm has.» on goo Hue-o. N xom uriv son“... .- _ x0. xn RMP$M v: 44 milliseconds. The LF351. while slower. recovers in nanoseconds from saturation. Difficulty with single beam instruments are variations Of the output DC voltage level owing to fluctuations over time (seconds) in power supplies of the PMT and the probe source. changes Of the output from the probe beam. component heating. sample distortion. etc. Conventional double beam spectrometers resolve this problem by using a mechanical chOpper and alternating between a signal and a reference. and correcting appropriately. In conventional single beam instruments the user has to set the OST (transmittance) and 100GT by hand before making a measurement. A single-beam kinetic spectrometer. once Operating. does not allow for readjustment Of the OST and 100hT before every flash. and analogue to digital (AID) converters have a finite voltage range over which they convert. For Optimal use of the digitizing range it is best to have the DC voltage level at 0.0 V. The detector amplifiers have a DC Offset control which is set at the beginning of a set Of flashes; however. the Occurrence of slow fluctuations in the DC level are not compensated for and can drift out of the digitization range. A sample-and-hold (S/H) amplifier (Analog Devices.SHAr5; NOrwood. MA) based automatic offset were developed (Figure 2.6) to eliminate these fluctuations. Ten milliseconds before the AID converter is triggered to begin converting the SIR amplifier is triggered into a ”hold" mode. The DC voltage level at that time is automatically subtracted out Of all subsequent signals until another pulse is sent to the SIR to return to a "sample" mode. Over time scales short in relation with the slow variations in the DC level. the resulting output signal from the SIR is 45 FIGURE 2.6 SAMPLE-and-HOLD AMPLIFIER (SIH). The other two amplifiers after the SHAr5 are LF351s (not labelled). The input line is divided with one lead going to the SHAr5 and into the first LF35l's inverting input. and the second lead ties up in a voltage summing configuration going into the scond LF351's inverting input. Ihen the SHArS is in the "SAMPLE" mode it passes the input signal completely. it's inverted and summed with the positive signal for an output signal Of 0.0V. Ihen the SHAr5 is triggered into the "HOLD" mode it stores the voltage on a capacitor and holds that voltage on its ouput line. it is inverted. and subtracted from the prsent signal resulting in an output signal centered around 0.0V D.C. 46 >9 2 >9- >9. >9- >9. >9- » w w w... E- EE- E. i. ._ .w 2 ii. EHWSL nern O)? >n¢h 47 an accurate representation of the input signal minus only the DC level. Absolute absorbances typically are not measured. I am usually interested only in the change of absorbance. AA. induced by excitation of the sample. Absorbance before excitation can be written as : . O A1 LOG(IiI11)- (2.1) where A1 is the absorbance; 12. the radiant intensity upon the sample; and Ii' the transmitted or unabsorbed radiation. Similarly. after excitation A. - 100(Iglif). (2.2) AA then is equal to the difference. Af-Ai. or. expressed in terms of light intensities: AA - Loc(Ig/If) - LOG(12IIi). (2.3) Rearranging. and assuming I? - I2. gives: 11 - Mai/If). (2.4) The output Of a PMT is proportional to the intensity Of light; under the definition of the voltage Offset. V(°ff) 'Ii' then AV is proportional to the change of light intensity. If-Ii: At the start Of a series Of flashes. the DC voltage is set to 0.0V and the Offset voltage is set equal to V(°ff). AV’ is just the difference from 0.0V induced by the laser flashes. 48 2.2.2 Timing Circuitry - This circuit is the control interface between the computer (discussed below) and the discrete components Of the FAS. Its layout is as in Figure 2.7. The computer provides four address lines for the decoder and a line which a ”timer” toggles "HI" (+5.0 V) and ”L0” (0.0 V) under software control. The rising and falling edges Of each pulse are shaped into pulses used to trigger a monostable to send out a 1.5 ns negative pulse (see Figure 2.8) which is connected to all the NAND gates. Meanwhile the decoder selects one of its output lines to send "LO”; the NAND gate with both inputs selected "LO" would produce a ”HI" output pulse for 1.5 us on the apprOpriate pin. A timing schematic is as shown in Fig 2.8. minimum time. as dictated by hardware. between pulses is 2.5 ps; maximum time is undefined. This interface has twelve lines going out. with room for four more if the current circuitry is used. and provides immediate and interactive control over any device requiring a standard TTL pulse. A cautionary note is in order here. If any Of the external devices develOps and reflects down its control line high voltage/frequency spikes. or if the control line acts as an antenna and picks up these spikes. this will affect the integrity of the interface. other devices. and the computer itself. The Phase-R laser is renowned for this as well as its RF (radio-frequency) interferences. and has been throughly grounded. shielded. and isolated as described as in (25). 49 FIGURE 2.7 TIMING CIRCUITY. The AIM-65 is required to bring four address lines. A0 - A3. and a timing line (labelled TTL). in to run the circuitry. The 74C154 is a 4 to 16 decoder and provides 16 discrete output lines (only 12 are currently utilized). The 7400 (NAND gate) and 7404 (HEX INVERTER) are used to shape each transitions on the TTL line into pulses (detailed description in text and logic involved in FIGURE 2.8). A monostable vibrator (74121) trims the pulse duration to 2.0 as. and the output is fed into input of each of 7402 (NAND gate). The NAND gate selected will be the one which has its input line from the decoder LO; when the other input. from the monostable. goes LO it will produce an output HI pulse of the same duration as the monostable's. 50 > (D 4- J1. TTL GNO 51 FIGURE 2.8 TIMING SCHEMATIC. The logic involved in transforming each transition (LO -) HI or HI -> L0) on the input TTL line. A. into a final output line. F. with one pulse per transition. The output Of each gate is labelled and is matched with a time trace below. L—F—L—— - I I 2 I 2 I c SIGNAL OUT 52 53 2.3 COMPUTER 2.3.1 General - Three computers are directly involved in the Operation of the FAS. and a fourth if extensive numerical computation or high precision is required. Immediate data storage is provided via a modified (discussed below) NICOLET 1074. Instrument control is provided through a ROCX'ELL AIM-65 micro-computer. Long-term mass storage and data processing is provided by one of two DEC (Digital Equipment Corporation) PDP-ll (either I02 or I23 processor) minicomputers . Each will be discussed in turn below. 2.3.2 Nicolet 1074 - The NICOLET 1074 (Fabri Tek. equipped with an SD-72I2A 9-bit AID converter and a S! 71A sweep controller; Madison. '1) is Of early to mid 1960's vintage. and contains four memory channels of 1.024 bytes (commonly referred to as "1R") where a byte represents a data point Of an arbitrary integer representation Of less than 29 and greater than -(29). The internal 9-bit AID has a dwell time Of 20 us per point for 1024. 2048. or 4096 bytes (or points) sequential in time. Mbdifications by C. Yerkes. B. Buttner (members of the Babcock Laboratory). and M. Rabb enabled direct addressing Of the NICOLET's data registers from an external source. This allows use of external AID. with dwell times between 60 ms to 10 us for a total Of 1023 points. to write into a selected memory channel of 54 the NICOLET. Other modifications include multiplexing ability which permits 4 sequential triggers to be loaded into the four memory channels. This is an ideal feature for experiments looking for periodicity Of less than or equal to four. Principal use Of the NICOLET is as a signal averager; repeated triggering causes the new data set to be summed with that already existing. It also contains the hardware to maintain a ”current status” Of the data contained in its memory on an oscilloscope. provide hard copies. scaling. and some elementary arithmetic processes. Duplication Of these features in the AIM is feasible. but the overall performance Of the AIM and the system as a whole would be degraded severely. 2.3.3 AIM-65 Microcomputer - Tending the timing device circuitry and the NICOLET is the AIM-65 microcomputer. which serves as essentially the ”brain" of the system. The AIM-65 is a single-board microcomputer which is extremely flexible owing to its design as a system development unit. The heart of the hardware is an 8 bit 6502 microprocessor unit (MPU) with a clock frequency Of 1MHz. The AIM. in its present configuration. has an 8! ROM (read only memory) ASSEMBLER. 8! 11011 BASIC interpreter. 4! RAM (random access memory) for programming. and a 4! ROM MONITOR and EDITOR. Memory expansion up to 64 X is aecomodated through its parallel expansion connector (not yet implemented) and device interfaces through its parallel applications connector. An onboard keyboard. 20 character display and 20 column thermal printer. and 2 PSI (frequency-shift keyed) cassette lines allow this MPU to be used in a truly "stand-alone" 55 environment. Use Of a cassette recorder as a mass-storage device was quickly discarded in favour Of using the POP-11. As the AIM provides for an Off-board use Of a terminal. a circuit (73) to convert the 20 mA current loop to a standard R8232 was built. and all communication tO the AIM is now accomplished through a Zenith Z-19 terminal and/or the POP-11. Tho parallel ports. A and B. Of 8 lines each with a variety of control lines for ”hand-shaking" reside in the applications port. Port B also is connected internally to a user-definable R6522 (Versatile Interface Adapter. Rockewll Int'l. Corp.; Anaheim. CA. FOr general usage and detailed explanation Of Operation see (74).). a integrated circuit chip containing two 16 bit timers which run at a clock frequency Of 1MHz. Timer 1 Of the 6522. PB7 (Port B line 7). can be software configured to invert the voltage on PB-7 each time it completes a countdown. An automatic reloading feature permits the user to set the timer running and then to load a new time into the latches of the timer without affecting the current timing setup. Completion Of the count inverts the voltage level. automatically loads the new time. and sets an interrupt. Four other lines Of Port B (PB-0 - PB-3) are used to select the correct addresses for the decoder in the timing circuit. Port A is used for data transfer from the NICOLET to the AIM and subsequently to the POP-11. The data is encoded in inverted binary coded decimal (BCD) logic in coming up from the NICOLET to the AIM. translated into ASCII code and is sent in a serial fashion to the POP-11. A full memory transfer from the NICOLET requires 90 seconds (the initial version Of the data uploader was written in BASIC and 56 required 20 minutes) at 9600 baud (9600 hits per second) The POP-11 is thus used as a mass-storage device during experimental runs. and later for data analysis. Initial analysis consists Of automatic file documentation which provides a safeguard for experimental parameters. and conversion Of initial raw voltage data into absorbance data. Each data file is fully compatible with the Chemistry department's CEMCOMGRAF facility as well as the more limited facility of the Babcock Laboratory; additionally each file is structured to be amenable to a more sophisticated data-base management protocol currently being develOped. Additional processing is available through a prototype data file handler developed during the course Of this work and discussed in more detail in the next Chapter. Fer high-precision computations or cases where numerical overflows are likely (PDP-ll computers are limited to -1o37 1nd 1037) a modem link to the University computer can be used to transfer data to take advantage of the higher precision and greater numerical range. CHAPTER 3 SOPT'ARE DESIGN 3.0 PROGRAMMING CONSIDERATIONS Microcomputers have languages available to them from the high level ones such as BASIC or FORTRAN. intermediate ones such as FORTH. and low-level machine language (Hereafter loosely referred to as: ASS"LY.). Many commercial instruments which allow user interaction and control through specific application programs provide some form of BASIC. For applications where program execution speed is critical or where memory limitations are severe have forced most control programs to be written in ASSEMBLY. BASIC as an interpreted language is far too slow. FORTRAN requires large amounts Of memory and/or mass storage. and compiled BASIC suffers from the same limitations as FORTRAN. For environments in which the microcomputer is not forced to function as a stand-alone system and can interact with a larger "host” system. the software development phase can be done on the host and the programs then transferred into the smaller microcomputer. This type of implementation is used extensively in the MSU's Chemistry Department. The interactive language largely used is FORTH which executes much faster than BASIC and occupies much less memory. FORTH is also advantageous in that execution speed critical code can be written in 57 58 ASSEMBLY in the body Of the main program. not requiring the programmer to go through the tedious task of ”linking" tOgether the various program segments. BASIC has been used in this particular implementation rather than FORTH for two reasons. Initially BASIC was the only high level language offered for the AIM-65; therefore. initial programming was done in this language. Second. more scientists are fluent in BASIC instead of FORTH; thus comprehension and modification of the program by others will be an easier task than if written in a relatively Obscure language. Another way to address the host—microcomputer interaction is through a ”master-slave" interaction. which is standard for most typical computer peripherals such as lineprinters or terminals. In this instance the peripheral. in this case a microcomputer. is given enough ability through hardware and permanently implanted programs in ROM to perform very specific tasks. Interaction with it is through the host computer only. An immediate effect of this is a reduction in the memory size the microcomputer programs require as terminal drivers and other input/output (I/O) routines are no longer required. Appendix A contains a listing Of the three routines with their inline documentation used in the AIM-65. 3.1 AIM-65 3.1.1 System Execution - At the expense of memory. the main routine has been written to give informative prompts and error messages as necessary. Upon loading and 59 running the system. the user is queried as to the function to be performed. new or Old timing setup. data uploading. or exiting. New timing setups reinitialize all memory cells used for data storage. The user is then prompted for the number of pulses to be sent. number of cycles. Fer each pulse the user is requested to enter a device name. time (in microseconds). and the address requested. FOr both Old and new timing setups the integrity of memory is checked. and error messages are generated and execution halted if needed. Operation Of the Kinetic Operating System can be through the POP-11 or as a stand-alone system. Only in the case of data uploading is actual linkage to the host computer required. The system is general and quite versatile in its timing functions. 3.1.1.1 Program Description - Three programs provide the necessary hardware timing control. data uploading. and interactive with the user. Program listings with in-code documentation are in Appendix A. Brief descriptions Of their functions follow. KOS.BAS is the BASIC routine which provides the user with a limited number Of Options. initialization of memory regions according to the Option selected. and interactive querying for necessary information to fulfil the Option. Most microcomputer BASIC interpreters allow for linkage to user written assembly routines. The requirements for this linkage involve placing into specific memory cells the "vector” or address pointing to the start of the program; then a function call similar to : I - USR(NUM). is executed. "NUM" is a floating point 60 number passed directly to the Assembly routine. Microcomputer development systems are limited in readily available utility programs (i.e.. there are none) for aiding software develOpment; thus. linking loaders (prOgrams which are capable of taking several separately develOped inter-dependent programs. resolving addressing differences. and loading them into memory as one program) are nonexistent. and program linkages are the user's responsibility. Modifications to assembly programs usually change starting addresses. sizes. etc.; therefore. an implementation which has been adopted here. to minimize the number of program(s) modifications each time a single program is modified. is to place at the top of memory the starting address Of variables and of the code itself. KOS is thus required only to PEEK (a function permitting direct examination of memory through BASIC) to find the starting address. place them in the JUMP handler. and then to execute a function similar to the above. This protocol requires fewer changes than otherwise. and suffices in the absence Of a relocatable assembler and linking loader. KOSTIM.AIM is an assembly routine which executes during the timing sequence under interrupt control. and directly controls the timing circuitry. The timer. the R6522. is only a 16-bit timer providing a maximum of 65536 us between completions (.066 seconds). To circumvent this restriction a third byte is kept in memory which KOSTIM decrements each time the timer completes its count. Ihen the third byte reaches zero. KOSTTM advances to the next devices's timing delays. loads them into the timer's latches. and continues as before. Three bytes for specifying a time gives a maximum Of 16.7 seconds between pulses. and 61 the increased overhead in program execution raises the minimum time between pulses to 150 us. 'hen the third byte is used the timer still times-out and toggles PB-7. but all possible device addresses will be de-selected except on the last time through the iteration. An earlier version printed out the number of iterations completed. but the AIM uses the second timer Of the 6522 for baud rate generation and a hardware interference exists between the two timers causing inaccurate timing sequences. At present data storage exists for 24 TTL.pulses to be sent out per cycle for 255 cycles. No restriction exists on which of the TTL outputs is selected nor how many times in a cycle. Transfer rates from the NICOLET through the AIM to the POP-11 is close to the maximum that an POP-11 floppy drive based system can handle. A recent upgrade to an RSX-llM multi-tasking/user Operating system cannot support this high transfer rate. KOSNIC.AIM is an Assembly program which uploads the data from the .NIOOLET and handles the output directly to the POP-11 itself. Memory constraints require that data transfer from the NICOLET occur on a number by number transfer Of nine bytes per number 3.2 LSI-III02-23 HALCDC.MAC (65) was initially written for use with a CYRERPVSO and allows the POP-11. Operating under an RTrll operating system. to function as a terminal emulator while retaining the ability to pass or receive data locally. Relatively extensive modification permits CDC to serve as the interface link between the PDP-ll and the AIM. Data transfer rate between the two is at 9600 band (960 characters per 62 second). The major detriment to this arrangement is the monopolization Of the POP-11 during the course Of the experiment due to the restrictions Of the RT‘ll operating system. Recent computer hardware upgrades allows use of RSX Operating systems; however. CDC will require further modifications. as of this writing. to function in a multi-user environment. Data conversion and manipulation is performed by KINSETtFTN and MINI.FTN. MINI converts data files uploaded. each of which consists Of only time sequential lists Of voltages. into files maintaining the initial rsw voltages. the corresponding time (in micro-seconds). and the change in relative absorbance as calculated by equation 2.5. MINI prompts for the necessary information and for all additional experimental parameters. performs the necessary conversions. and stores the data in an unformatted binary sequential file. All Of the records containing parameters of the experiment are "tagged” with "DO". and all data records with "RD." This file serves as a"unit" for a future data base implementation and helps to insure that experimental parameters will be retained. The file created is compatible with MULPLT. the Chemistry Dept.'s Graphics Facility maintained by Dr. Ten V. Atkinson. as well as the more limited graphics facility which has been implemented during the course of this work on the Babcock laboratory's POP-11. KINSET.FTN is a generalized file handling program developed here. It's initial implementation in a PDP-11/02 RTrll system was limited to performing arithmetic operations between two files and some manipulation of the data within a file. A recent upgrade to the POP-11I23 RSI-11M system takes advantage of virtual array space and faster disk access 63 times. It's Operations are on or between fields Of data. where each field is a complete set of x.y data (an inclusion for individual weighting Of each datum will be in the next upgrade). Operations on a field include base-line corrections by variable order polynomial fitting. multiply. divide. add. and subtract functions. numerical differentiation and integration. and various smoothing algorithms. Operations between data fields are limited to the four basic arithmetic functions. Iindowing. terminal plotting. listing facilities. and several file output formats are available. This program. in effect. eliminates the need for most specialized data manipulation and conversion routines typically found in scientific environments. Dynamic display and handling Of large quantities Of data can give one a "feel" for the data. and better insight into solving particular experimental or model problems. The next version will include Fourier and Inverse Fourier transforms. digital filter techniques for enhancing signal to noise ratios. spline fitting algorithms for interpolation and extrapolation. and an increase in data base size. A fourth program. CFIT.FTN is a completely general non-linear parametric-fitting program implemented for kinetic analysis on a POP-11 computer system. CFT4A.FTN (35). a subroutine. has been modified to serve as the "algorithm" unit for CFIT. Ihile CFIT is useful for small sets Of data (less than 100 points). hours can be required when fitting a set containing 100 points; therefore. its overall usefulness is 64 limited owing to a relatively inefficient "minimum-seeking" algorithm in CFT4A. and slow processor cycle times Of the PDP-ll compared to the CIDER-7 50 . CHAPTER 4 EXPERIMENTS AND FUTURE PLANS 4.0 MATERIALS and METHODS 4.0.1 Chloroplast Preparation - Class II chlorOplasts were used in all experiments described in the following sections. and were prepared as according to Robinson et al. (23) and briefly summarized here. Commercially available spinach was washed and depetiolated (leaves separated from the stem). The leaves. 500-1000 grams.. were ground in "Grinding Buffer”. a solution of 0.4 ‘! NaCl. 2 m! MgClz. 1.0 mgIml Bovine Serum Albumin (BSA). 1 m! Ethylenediamine tetraacetic acid (EDTA). and 20 m! HEPES at a pH of 7.6. for 10 sec in a Faring Blender. The homogenate was then passed through 8 layers of cheesecloth; the supernatant centrifuged until 3500 x g was reached and then stOpped. The supernatant was then resuspended in the "First Resuspension Buffer". a solution of 150 m! NaCl. 4 m! MgClz. and 20 m! HEPES at a solution pH Of 7.6. and centrifuged at 7500 x g for 10 min. After centrifugation the pellet was resuspended and stored in SHN. a solution of 0.4 ! Sucrose. 10 m! NaCl. 50 m! HEPES at a solution pH Of 7.6. ChlorOplasts were stored typically at -40°C with no apparent loss of photosynthetic activity; however. the question of membrane integrity 65 66 is taken up later. Final chlorOplast concentration as function Of chlorophyll concentration was determined as by Sun and Sauer (24). ChlorOplasts were diluted 1:200 into a solution Of 80$ acetone/205 H20. filtered. and absorption at 652 nm measured in a 1 on cell. Concentration in.mg ChlIml was determined by multiplying absorbance by 5.8. Photosynthetic activity and integrity were assayed at time of preparation and before each experiment by measuring steady state rate Of oxygen evolution. Method and apparatus were developed in the Babcock laboratory by Yerkes (25) and Buttner (26). ChlorOplasts were diluted to 70 ugIml in a solution Of 50 m! NaCl. 2 m! MgClz. 20 m! HEPES. and 3.6 m! K3Fe(CN)6 (an electron acceptor) at a solution pH of 7.5. Equilibration for 30 sec preceeded the actual measurement. Two sets of measurements were always made; the first as above measuring the ”coupled" rate Of oxygen evolution‘l) with the addition of methylamine. cn3sf2' to a concentration of 10 m!. as an uncoupler. measuring uncoupled oxygen evolution. Uncoupling usually stimulates the rates by three to four fold. Figure 4.1 shows typical traces Of coupled (4.1A) and uncoupled (4.1B) rates over the same period of time. This procedure provides a crude estimate Of the intactness of the entire photosynthetic "1)"In this context coupled oxygen evolution means that the entire photosynthetic chain from the OEC to the ATPase is Operational. Water is being oxidized and ATP is being produced. Uncoupling permits the OEC to continue to oxidize water. but ATP is no longer produced. 67 FIGURE 4.1 OXYGEN EVOLUTION. This depicts typical oxygen evolving rates seen with class II chlorOplasts prepared as described. These rates were Obtained with freshly prepared chloroplasts. Rates obtained with freeze-thawed chlorOplasts (not shown) are similar. (A) Coupled oxygen evolving rate. (B) Uncoupled oxygen evolving rate. 68 u _ _ 2.209.: 03.2.. 2 1min 69 apparatus from the OEC to the ATPase. Only chloroplasts with uncoupled rates of greater than 200 uncles Ozlmg Chl-hr were used. TRIS-washed chloroplasts. while not used in any of the experiments described herein. are planned for some experiments and are used in many of the experiments described in the literature. and are thus described here. TRIS-washing is perhaps the mildest known treatment for inhibiting oxygen evolution. The preparation of TRIS-washed chlorOplasts is as above except for two additional suspension and centrifugation steps. After the second centrifugation (described previously) the chloroplasts were suspended in a solution of 0.8 M TRIS.and 1.0 m! EDTA at a solution pH of 8.0. and incubated under room light at 4°C for 20 minutes. Afterwards they were centrifuged at 4000 x g for 10 min.. the pellet resuspended in the first resuspension buffer. centrifuged at 4000 x g for another 10 min.. and finally suspended and stored in SHN as described above. 4.0.2 Standard Experimental Conditions - Standard class II chloroplasts. evolving greater than 200 umcle Ozlmg Chl-hr. were used at a typical chlorOphyll concentration of 49 u! in a total solution volume Of 1.5 m1. A quartz cell with 0.5 cm path length held the sample positioned before the PMT in the path Of the probe beam. Monochromator bandpass was usually 4 nm. PMT voltage was between 530 - 600V. Recalling Figure 1.1. NAD'P+ is the final electron acceptor in vectorial electron transport through the two photosystems. Chloroplast preparation. however. disrupts the ferredoxin-NADPH reductase complex 70 which interfaces the one-electron chemistry Of ferredoxin with the two-electron reduction of NADP+. This entails the need to add exogenous electron acceptors to the reaction media. Fer experiments requiring operation of both photosystems Methyl Viclogen (MV) is often used as it removes the electrons from PSI. For experiments involving only PSII DCMU and ferricyanide can both be used together. DCMU blocks electron transfer between 0 and Oh. and prevents PSI from Operating. a Ferricyanide accepts electrons then directly from 0.. albeit slowly. Reaction media was typically SHN except for experiments involving proton release/uptake measurements. Reaction media experiments involving proton measurements consisted of 10 m! KCl. 2 m! MgClz. and (if present) 1.0 mgIml BSA at a solution pH of 7.6. Electron acceptors were as described (above). A typical timing sequence is depicted in Figure 4.2. Generally. the sample was exposed to the probe beam 50 ms before triggering the SIR. AID. and Laser to allow settling time for the PMT. Immediately after conversion ends the SIR reverts to "sample [" mode and the shutter is closed to reduce perturbational effects from the probe beam. Tetal exposure time throughout the entire experiment was no more than 1.0 sec. 4.1 CONTROL EXPERIMENTS 4.1.1 Steady state effect of Probe Beam - Figure 4.3A shows the effect of the probe beam continuously illuminating a standard sample in which both photosystems were 71 FIGURE 4.2 A typical sequence of timing events (not to scale). On the tap line. A. each ti indicates the beginning of a timing cycle. The cycle repeats itself at every ti. the smaller line after the t1 marker indicates that the active event timing is comlpleted. and the AIM-65 enters an idle lOOp until the beginning of the next cycle. Line B is an expansion of one cycle. and line C is an expansion of the active event timing portion. Example times are given below (note: all times are given with respect to the previous event) : time dpration gygnt tia 50ms photoshutter open tib 10‘. SIB -’ 'hOld' tic 0-1ms AID - begin converting tid 1-20ms NICOLET - begin storage tie 10ms SIH - 'sample' tlf 1ms photoshutter close tlg 1-15s enter idle lOOp 72 on. K. on. as. ou. an. S... V E _ _ _ _ _ _ . . . _ . a .A 9:53 aco>c o>$oc V. l f f f l f l 0 o 4 4 4 4 4 4 4 . l 0 0 O l o 4 4 4 4 4 . 4 4 4 4 4 4 . 44444 . 0 f f l l c 4 4 4 . c u r _ _ F as; 22 an: 0.] a.’ e . . o . 73 FIGURE 4.3 PROBE LIGHT EFFECT. The effect of contiuous illumination Of the sample by the probe for: (A) 40.92 seconds. and (B) 1.0 second. 74 OJIZ F 0.056 - 352824 I 20.5 Time (seconds) o>=2mm ouconbmad o>=2mm 75 Operational for 40.9 sec. and Figure 4.38 is the same for 1.0 sec. An overall decrease in absorbance is seen over 40.92 sec; however. photochemistry induced by the probe beam is expected to increase the absorption. Clearly. what is happening is that the sample is settling and decreasing the absorption. Figure 4.38 reveals no significant change in absorption. The noise level is high as the time constant for this particular trace was less than 10 us. 4.1.2 Flash-induced Difference Absorption Spectrum - As discussed in Chapter One and depicted in Figure 1.5. the effect of an absorption band shiftis the derivative of the absorption band. In ChlorOplasts this shift is observed from 400-570nm. Figure 4.4 is a plot of a flash-induced electrochromic difference spectrum and follows the expected derivative shape. Two measurements. each consisting of 25 flashes spaced 10 seconds apart. were made to obtain each point. Both measurements were taken under standard conditions. excepth that the second included 5 u! Gramicidin. The second set of data would then contain the same set of absorption changes as the first except for those due to the electrochromic effect. Subtraction of the two sets of data would according to Equation (4.1): eqtabl eqon A(AA) - '(AAstd - AAgrm” (4.1) eqoff where A(AA) is the difference in relative absorbancies; AAstd the induced difference with no chemical modifications of the electron 76 transport chain; and AAgrm' the induced difference with Gramicidin to eliminate the electrochromic responses would yield a flash-induced electrochromic 77 FIGURE 4.4 FLASH INDUCED DIFFERENCE ABSORPTION SPECTRUM. Center line represents zero absorption changes. Each "+" marks one point. 78 .nmm A85 1525935 .09. _ NOIMNm TOI Nolmmm—d 30NV88088V BAIIV'BH 79 spectrum. As discussed. in Chapter One. the electrochromic response is a result Of the electric field set up by the charge separation across the membrane. Gramicidin in nanomolar and above quantities causes the membranes to become ion permeable. and allows rapid decay of ion or electrical gradients through high ion fluxes. This results in almost immediate collapse of the charge separation and. rapid decay of absorbance changes due to this effect. 4.1.3 Effect of DCMU on the 518 nm change - As discussed previously in this chapter. if ferricyanide is used as an electron acceptor and DCMU as a blocker between 0. and 0b then only PSII is Operational. the result of this is that only PSII contributes to the trans-membrane charge. i.e. only one half of the normal charge should be generated across the membrane. The electrochromic absorption change at 518nm should then be one half Of that normally seen. Figure 4.5 reveals this to be so. Tho traces with their baselines superimposed at the time of the flash. are in Figure 4.5. both were taken under standard conditons. 25 flashes each with 10 seconds between flashes. Trace B has had 166 u! DCMU added. Higher concentrations of DCMU did not decrease the response in DCMU-poisoned chlorOplasts. Lower concentrations of DCMU lessened the effect. The amplitude of trace A is double that of trace B. This is in agreements with Hong et al.(55). Schliephake et al.(56). and Joliet et al.(77). 80 FIGURE 4.5 EFFECT OF DCMU ON 518 CHANGE. Baselines of the two traces shown were overlapped to emphasize the difference in absorption changes. Trace A is a normal 518 change. and trace B is the response with 166 u; DCMU. The negative spike(s) are at time zero. and is the PMT's response to the laser excitation. Time constant was less than 5 9 10.6 sec. [DCMUJ-166 u!. (MV)-49 RM. and equilibration for 60 seconds was allowed. Relative Absorbonce 81 0.1 21 85 '1 - A -. 3 -0.05252 - 1 P 4.22000 4 ,1 i 4 -200.00 44100.00 10000.00 Time (micro-seconds) 82 4.1.4 Neutral Red Absorbance Changes - Operation of photochemistry pumps protons to the interior of the thylakoids in the oxidation of the PO pool. Protons are also released to the interior by water oxidation. These protons are used by the ATPase in the production of ATP (see Figure 1.3). Studies performed in which Cresol Red. a non-permeant indicator dye. was used to show that external proton release and uptake can be time resolved (milliseconds or more; for a detailed discussion. see Junge (2)). Neutral Red (NR). a membrane permeable indicator dye. can respond to both external and internal proton release or uptake. Selective buffering of the external media by BSA. will allow only the response of NR to internal proton release or uptake to be observed. Figure 4.6 reveals the response Of NR at 529 um. Proton release is indicated by the increase in absorbance. The time scale of 40 ms is too short to show the eventual leveling off and decay in absorbance (0.5-l.0s). Two measurements were performed to obtain the trace. A standard sample with no NR.present was run in which 25 flashes spaced 10 seconds apart were averaged. A second experiment was carried out under the some conditions with 40 u! NR. The former was then subtracted from the latter to obtain the trace in Figure 4.6. Correlating the NR response to particular phenomena. such as the Sn state transition(s) responsible for proton release. has led to conflicting ideas as discussed in Chapter One. Experimentally. the controls Auslander and Junge established (22) are adequate. Hewever. the integrity of the chloroplasts used and the reproducibility Of their condition are suspect. It has been demonstrated that freshly prepared 83 FIGURE 4.6 NEUTRAL RED ABSORBANCE CHANGES. As discussed in the text. this is the result Of subtracting the absorption changes at 529nm with 40 A! N.R. with those obtained without N.R. in solution. Both original data sets contained sharp negative peaks at time zero. subtraction took most Of it out. Each data set also contained laser-induced scatter. Subtraction of one from another automatically eliminates the scatter. Relative Absorbance 84 0.5355-03 'T l l l l -lr- l -o.394£-03 ,L i 1 -500.00 2250.00 5000.00 Time (micro-seconds) 85 chlorOplasts. those stored at -40°C. and those under liquid nitrogen all show different proton release patterns as detected by NR. Freeze-thawed chlorOplasts have membranes which are permeable to buffers such as HEPES. MES. and phosphate. Freshly prepared chloroplasts do not display this dependence on buffers. Both sets of chlorOplasts continue to display otherwise normal photosynthetic activity. These experiments show that in general the FAS is able to reproduce results in the literature concerning electrochromic band shifts. inhibitor effects. and proton release as measured by an indicator dye. The smallest significant absorbance change (see Figure 4.6) is 2.3 ° 10'4 with a signal-to-noise per flash of 0.4. Bewever there are experimentally some problems to resolve. Figures 4.5 and 4.6 reveal fast negative absorbance changes just after the laser flash. This is due principally to the failure of the cutoff filters to attenuate the laser line fully. Essentially. this introduces a ”dead-time" Of lOOus before useful information can be extracted. Fer sampling rates of up to lOns per point (200kHz; 10ms for a full scan). subtraction of a separate data set containing only changes due to scattering gives reproducible results. Higher sampling rates (ranging from .06-2 us per point or 16.6 - .5Mhz) are limited though. A gating technique was tried which turns the PMT Off for a few microseconds. (The method and equipment for this technique were develOped by Buttner (75)). This entails making one of the first few dynodes in the dynode chain in the PMT highly negative with respect to the photocathode. thus preventing any electrons from leaving the photocathodic surface. During the duration of the gate pulse the laser 86 would be triggered. and. presumably. the PMT would be able to track after the gate was over the intensity changes as it would not have been affected by the laser flash. However. under continuous photon flux upon the photocathodic surface (from the probe beam). and while the gate is applied. electron carriers accumulate until the gate is is removed and a large electron flux then travels down the dynode chain. The immediate effect is an apparent large absorbance decrease upon turning the gate off lasting for hundreds Of microseconds. 'hile a laser scatter artifact is no longer observed the apparent absorbance decrease is of a larger magnitude and of a longer duration that attributed to the laser artifact. As a result this technique is no longer pursued. The laser scatter difficulty is being resolved through a redesign Of the sample enclosure as mentioned in Chapter Two. Placement of the sample compartment closer to the PMT. restriction of the light path from the cell to the PMT window. choosing filters of higher attenuation at 640nm and lower attenuation at lower wavelengths. and ttenuation of the laser beam itself should remove the laser scatter. An increase in the signal-to-noise ratio is being affected by a redesign of the sample cell itself. The cell pathlength is begin decreased to 2.0mm from 5.0mm. and all the surfaces not directly involved or needed in the transmission of either the probe or excitation beams will be black-masked to reduce scattered light. The above has already been partially completed. and the tests appear promising. The primary purpose of this project. the construction of a FAS. has been achieved and experiments of interest are being 87 devised. The modularity of the FAS is a tremendous advantage over commercial designs. Instrument modifications will be relatively simple. requiring only for the affected module to be replaced and/or modified. The computer interface is quite flexible. and literally requires only a change Of software to accomodate differing experimental requirements. 4.2 PLANNED EXPERIMENTS 4.2.1 Spectral and Kinetic Identification of Z - In the literature today there is evident much interest in the primary electron donor to P680. Since 1976 there has been inferred a relationship between Signals IIvf and IIf and 2 (49.50) has been inferred. Bocques-Bocquet (9). to explain the nano- and micro-second reduction times of P680+. has invoked the idea of two donors acting in parallel. Renger. using controversial "inside-out" TRIS-washed chloroplasts (52) and brcmo-cresol purple. a nonrpermeant indicator dye. quantified proton release thought tO be related to the redox reactions of Z and correlated it to the extent of P680 oxidation. O'Malley (11). has suggested that Z is a quinone. and in its oxidized state is a cation radical. Z'+. giving rise to an ESR detectable signal. Medal studies he has performed on quinone cation radicals generated in 12 3.32501 give remarkably similar ESR lineshapes to that of Signal II. Diner. at the 6th International Photosynthetic Congress. provided evidence Of an Optically detectable signal around 320nm seemingly correlated to P680+ reduction. Other closely related components of PSII are also yielding information. The amino acid sequence of the P680 containing protein has apparently been determined by Trebst (19). Empirical models determining protein folding. regions of hydrophobicity. and hydrophilicity support general configurational models such as Johnson's et al. (62) for the OEC complex and its interactions with Z and P680. Dekker et al. (76) have apparently obtained Optically the difference spectrum of Z-f-Z. and kinetic data related to the reduction of Ze+ in the UV region. Use Of either a deuterium lamp or a xenon arc lamp should permit us to do likewise. An important set of experiments will be the correlation of the behaviour Signal II with that of Z. Another set of studies will place emphasis on the extraction/reconstitution of Z. This is important because while a structure for Z has been proposed this can only serve as an unproved model until the component can be isolated and identified. Medification of ring substituents in Z-analogues. assuming Z is a quinone species. should yield much information regarding the requirements for electron transfer from the OEC through Z and into P680. HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH 89 APPENDIX A INTERACTIVE KINETIC OPERATING SYSTEM VERSION 17-JUN-83 Dwight H. Lillie Babcock Research Laboratory Dept. of Chemistry Michigan State University This is the interactive and initialization routine for the AIM-65 microcomputer. The total Operating system enables the AIM-65 to perform very fast timing sequences or to upload date from a NICOLET-1072 into the PDPbll. This version. as all previous versions. suffer from a lack of raw memory in the AIM-65. Thus code is condensed as much as possible. and this special documentation convention adopted for use with HALCDC. 'There are two main assembly language subroutines which control the actual perfonmance of the above functions. Linkage to these routines currently is implemented through peeking at pro-defined high locations in memory. Fer current implementation I've located the pointers as such: ‘TIMING SUBBOUTINE : SFFC.8FFD (low/high byte) (4092,4093) NICOLET can? SUBR : SFFA.8FFB (low/high byte) (4090,4091) This implementation has some advantages - namely in that fewer changes need to be made throughout the BASIC and ASSEHRLER code when modifications are (will be) made. At a later date when the AIM-65 is expanded to 64K these pointers will be relocated - preferably to zero-page; however. as BASIC keeps a number of pointers there - probably just under the SYSTEM routines. 90 dimension the arrays to SI 8 24 (for 4k of RAM) SI - 24 0144 DE(SI).NAS(SI).AD(SI) define a function which will yield the floating point number of the addition of two bytes. These are useful for determing the address of a location store in two bytes in memory. ADDR is address of the low byte and ADDR+I is address of the high byte: FNFP(ADDR)'INTIPEEK(ADUR))+INTIPEER(ADDR+1)‘256) 0411? menu-murmur))+m'r(1>mu+1)9256) 113-" ~ :ponz 40962.16 O0.000.000....OOOOOOOOOOOOOOOO find out what user wants to do eeeeeeeeeeeeeeeeeeeeeseeeeeeee RK-OzGOSUB 800 QHHHQMHHHHHHHHHHHe—HWHHH 10 PRINT'INPUT FUNCTION ::":PRINT" "."1 - TIMING...REPEAT" 20 PRINT" "."2 - TIMING...NE'”:PRINT" "."3 - DUMPING" 30 PRINT" "."4 - EXIT":INPUT J' 40 IF J84 THEN STOP IF J-3THEN1000 IFJ‘IANDn'l ms 85 nau- 00 if j-2 then we just naturally fall through to the correct spot or if j-l and we've never been here before - as indicated by kk this section devoted to setting up parameters for performing a timing sequence BA - base for timing arrays in.memory ND - number of device pulses per cycle also initialize pointer to start of actual routine O G nmaqzuqmmmn) initialize entire array [0 0 FOR I'OT0100:PORE BAiI.O:IF PEEK(HA+I)<>OTHEN9000zNEXT HHHHHHe—HHHHHHHHHHHHHHHH prompt to get the necessary input 9.1 [ in order to set up a timing situation I I 160 mmmxmam'rmn : “.113 180 mm'mszs m CYCLE : "sun 190 INPUT" on ITERATIONS : ";N1 200 K-BA+(ND-l) 210 FOR 1-0 '10 ND-l 220 PRINT'PULSE nn+1 :nm 230 mum . DELAY . PIN :";NAS(I+1).DE(I+1).AD(I+1) 260 IF w(1+1)-s 1mm AD(I+1)-l2 270 1-1:oosun 650 I parameters are stored in arrays: (overflow byte for time>(address>HB THEN 9010 320 IF PEER(K+SI92)()LB OR PEEI(K+SI)<)AD(I+1) THEN 9010 330 PRINT :PRINT :K-K-l 340 NEXT 350 150 :GOSUB 650 I [ put the number of iterations in (HB+1.LB). and the number [ of iterations (ND). I These live at the fixed offset Of the BASE of device array I + size (SI) of array94 + 1 - if number of devices [ + 2 - if number of iterations I 360 PURE BAiSI94+2.(HB+1):PORE BAiSI94+3.LB:PORE BAiSI94.ND I I clear the screen and print the current timing setup I and then query for "go-ahead" I ass GOSUB 800 390 mm'...mrrnuzmou comma..." 420 mm mam 430 11111141911113 : an: 440 PRINT :inmxms : ".141) 450 PRINT :11an "muons : ";NI 460 mm :PRINT'NAME”. mmrwannuss" 470 mm 430 FOR 1-1 '10 umpxm NASH).-.DE(I).AD(I):NE1T:PRINT:PRINT 520 IF n00 mm 550 530 I 92 INPUT"READY...(Y'1IN-0) ":J:IF 130 mm 385 I [ look at the starting point and poke them into the I jump routine and then go for it I I 550 POKE 4.PEEK(4094):POKE5.PEEK(4095):I-USR(0):POKE40962.16 560 GOSUB 800:PRINT"EXPERIMENT COMPLETED":KK-1:G0'1010 I I [ this routine returns the 999rea1999 time input as three [ bytes and which are then poked into the proper places for [ the timing code to find I I 650 TE-NI:IF.J)0 THEN TEPDE(I+1) 670 OB-INTTTEI65536):HB‘INT((TEPOB965536)I256) 690 lB-mT(TE-(B965536-HB9256):IF J-O THEN 740 700 IF HB>O 0R [BI->150 mm 740:1.8-150 720 PRINT'MIN. LB DELAY TIME ADJUSTED TO 150..." 740 RETURN this routine will clear the terminal screen (HEATH ESCAPE CHARACTERS!) and return the cursor to the top left of the screen PRINT’CHRS(27);(HRS(72):mRS(27);CHRS(74):RETURN NINLET READOUT this section just calls the NICOLET UPLOAOER. 'e peek into previously defined areas ct'mmmory to get the jump location and poke them intO the USR jump handler. NOTE: sometimes the NICOLET fails to initialize itself prOperly at the start - if so. it can be forced into initialization by poking a value in. HHHHHHHHHHHHHHGHHHHHHH 1000 INPUT"1 - GO AHEAD/0 -RETURN : ';J:IF J-O THEN 10 1010 POKE4.PEEK(4090):POKE5.PEEK(4091):POKE 40962.16 1030 IFPEEK(40960)(>175ORPEEK(40960)(>191THENPOXE40960.191:I-USR(0) 1040 GOTOlOOO I I [ ERROR ESSAGES - few as they are I I this primarily seems to be a waste of memory space; 9000 9010 93 however. it does insure that timing parameters are CORRECT! PRINT"ARRAY INITIALIZATION FAILURE":STOP PRINT"VARIABLE TEST FAILURE”:STOP 94 I I TIMING CONTROL I I I VERSION: 6.17.83 [ This routine directs a simple timing algorithm for use [ with the specific hardware built for the AIM—65 micro— [ computer. This routine is called from a BASIC controller [ and is already initialized and parametically set up by I same. I I I HOOKUP: [ memory locations : FFE-FFF(16) - LB/HB OF where BASIC [ routine has to jump to [ start [ FFC-FFD(16) - where assembly begins I I [ Dwight H. Lillie I l7-JUN-83 [ Babcock Research Lab I Dept. of Chemistry [ Michigan State Univ. I . I I 9-3280 ASEHRL SIZE I24 I I allocate size of blocks - 24 bytes each I specify where assembled code starts I OBASE 9-9+SIZE DBASE 9-9+SIZE TLBASE 9-9+SIZE THBASE 9-9+SIZE NUMBER «9+1 mo'r 0-0 +1 rrmr «9+2 mm 9-9+2 moons 44001 DDRA 41003 rum: 41004 ACR 441003 PCR -iA00c IER 4110043 nov -iA400 [ I [ real code begins here...as you [ can see - not too impressive MHHHe-ae-H—ue-eF-u—e-ee-ne-Ie-e HRH "HRH!- 95 for the time being disable possiblity Of being interupted initialize the VIA store the address of the interupt service routine to the interupt bin-116: .l 130v define ports as being all output set up the auxiliary control register -ACR load a dummy time into the timer (T1) just to get things started define the type of interupts allowed and enable the interupts TART SEI 1.011 300 STA PCR STA INOT TAX TAY CLD load the interupt vector and store to handler LDA (ISR STA IROV LDA )ISR STA IRQV+1 get the of iterations as passed from the program and store in a temporary counter LDA TTERATel STA TEMP1+1 LDA ITERAT STA TEMPl LDA $11111111 STA DORA STA DDRArl LDA $11000000 STA ACR to get things going give the timer a dummy time to count through with all address lines deselected. This gives us a chance to get our act together and find ourselfs. 1m in? STA TLATCH STA. TLATCR+1 IDA. $00111111 enable interupts only from timer 1 STA IER LDA $11000000 STA IER enable processor to be interrupted CLI BEBLP LOX STA STA STA CHECK BED DEC LDY BEG STA STA MATT BED DEC DEY BNE ENE DEC LDA STA DEC ENE PUAL BEG SEI STA STA STA STA STA STA ISR STA STA LDA NUMBER LDA DBASEr1.X DECODE TLBASE-1.X TLATCH+2 THBASE+1.X TLATCR+3 LDA INOT CHECK INOT OBASEPl-X CONT in? THBASE-1.X LDA INOT MATT INOT IAIT DEX ITER TEMP1+1 BEGLP ITERA141 TEMP1+1 TEMPl BEGLP INOT PUAL $00 INOT ACR TLATCH TLATCR+1 TLATCR+2 TLATCH+3 96 deselect all possible address lines for now 81712 0410000 set interupt flag to let world know 301 INOT clear interupt flag by reading low latch of TLATCH and go home timer 97 RTI I I [ now allocate memory at a specified [ place so that BASIC can hook up [ without extensive modification I I Came [ assembly of entire program starts here .IORD ASEMBL [ actual executable code starts here .IORD START HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH 98 NICOLET READOUT This subroutine is initiated from BASIC and directly handles the data uploading from the NICOLET 1074 to the POP-11. This subroutine represents a reduction from 15 minutes for a 4K memory dump to only about 3 to 4. SYSTEM SUBROUTINES: OUTTTY : EEA8(16) - handles the actual output to the terminal requires character to be sent to be in the Ace. HOORUP : memory locations : FF8-FF9(16) - LBIHB of where assembly begins FFArFFB(16) - LBIHB of where BASIC jumps to for beginning. MODIFICATIONS: 23-AUG-83 included a "termination" sequence at the conclusion of a nicolet dump. This aborts any "GET" in progress. and has no effect if one is not. Designed to work with HALCDC current memory requirments: 276 bytes (23-aug-83) 27-IUN-83 Dwight H. Lillie Babcock Research Lab Dept. Of Chemistry Michigan State Univ. initialization‘_declaration 9-3000 ASEMBL [ character storage from nicolet DIGIT 9-9+7 PERIOD .BYTE '.' or .BYTE J15 LF .BYTE J12 INOT 9-9+1 DFLAG 9-9+1 .MINUS PLUS I I I I TERM 99 .BYTE '-' .BYTE '+' encode the message in reverse to make it easier to print .BYTE J12.J15.'R'.Jl2.J15.112.J15 .BYTE '0'.Jiz.lis.'(' :35" "HF" IDA system allocations - $1000 - tense - #0900 this is where BASIC must jump for successful transfer Of control 800 DPLAG define portB as all input except for bit 5 500010000 PORTB+2 $07 $00 DIGIT.1 .X LOOP inn initially - let's give the "print” sign a bit of time to go high before returning to BASIC PORTS $10000000 PC1V REGST if not high after 255 micro - go home DONE now wait for PCl to go high PORTB i00100000 PCl' well. it's time to start getting real data in now - read the sign first- PORTB FOR AND CMP ENE STA JMP CMP EEO CONT1 STA EFETCH LOX FETCH JSR ISR EOR ORA STA ISR ISR HHHHHH CPX ENE arms-1H JSR LDX 100 in $00001111 $01 if not a minus then go on and test to make sure NEXT MINUS DIGIT BFETCH $00 noun $01 PLUS DIGIT 8 time to get rest of character - update pointer 01 so...send a fetch.wait. and then read character SENDH MATT PORTB remember - the NICOLET'has inverted logic so invert the bits 8171? and as the character is BCD mask off top 4 $00001111 and convert to an ASCII character for presentation $00110000 DIGIT.X CKPB7 and prepare to send another fetch SENDL in case our internal count is not accurate or we're not synchronous with the NICOLET - we monitor whether the NICOLET is telling us that's all the characters for that number INOT OUTLP reached end of the buffer for character yet? 807 FETCH if so then send a fetch so that NICOLET has time to finish decoding the next character and then print out the current character SENDH 401 24-94-4 LDA ISR INX CPX ECC ISR JSR IMP ge-H—H-IHe-I LDA STA "HRH IER ISR DEX ENE HHHHHH BAGA ”He-IMF! 101 number Of characters put out is longer than that read in as we have to send DIGIT-1.X OUTTTY 300 um if we're all done then re-inforce the earlier fetch and go for a nice long wait SENDH LIATT SIGN DONE it appears that we're all done dumping the data so clean up and return back to (sigh) BASIC $01 DFLAG send out a termination message to ”CDC" and return to interactive BASIC 11 ISR L'ATT TERM-1.X OUTTTY DLP IATT do all of our waiting here. recall that the AIM can run faster than the NICOLET canl SFF BAGA on just a normal IAIT should be roughly 40-100 micro-seconds and that's probably long enough SENDH‘_SENOL this is for bit 5 of portB and tells the NICOLET when we're doing another fetch 3:955:25 ESEEEgg 5 102 PORTB $00010000 PORTB PORTB $111011ll PORTB CKPB7 check on the status of PB-7 the NICOLET doesn't seem to tend this well so make sure it really is null before drawing any hasty conclusions 101 $041 ENDCHR PORTB $01000000 ONE 800 INOT $01 INOT now set up the already defined linkage for the AIM-65. assembly begins here ASEMBL and the BASIC must jump to here START 103 LIST OF REFERENCES 10. 11. 104 LIST OF REFERENCES Norrish and Porter (1949). Proc, Roy. SocH Ser, ‘A 200. Junge. I. (1976). Chemistry gag Biochemigtry 2f Plgnt Pigments 191.1; Chapt 22. Goodwin. T.'.. ed.. Academic Press. NY. Mathis (1977). Primary Proggssgs 21 Photosynthesis 269-302. ed.. Barber. J. Van Best. J.A. and Mathis. P. (1979). Photochgm, PhotobioI 31. 89-92. Van Best. J.A. and Mathis. P. (1978). Rgv, Sgi, InstrI 49 (9). 1332-1334. Junge. w. (1931). Bioghim, 0162b”, m 141-152. Crofts. A. (1974). Eigehim, Biophys. Act;‘ 357. 78-88. Ingle. J.D.. Crouch. S.R. (1971). Agglg Chgm, 43. 1331-1334. Bouges-Bocquet. B. (1980). Bioghim,I Bigphys, Agtg. 594. 85-103. Bernstein. M.. Rothberg. L.J.. Peters. K.S. (1982). 92:3. Phys, LottI 91. 4. 315-319. O'Malley. P. and Babcock. J.. iochig, Bigphys, A555; (submitted). 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 105 Bassham. J.A.. and Calvin. M. (1957). 1!; Path 2; Carbon 1; Photogynthgsis 1-104. Prentice Hall. Englewoood Cliffs. NJ. Hill. R.. and Bendall. F. (1960). Naturg 1.86.136. Hind. R.. and Olson. J.M. (1968). AnnI Rgv, Plant, Physiol. 19. 249. Avron. M. (1975). fiioenergggicg'gf Photosynthgsis 374-387. Govindjee. R.. ed.. Academic Press. Inc. New York. NY. Clayton. R.K. (1980). Photogygthegis: Phygical Meghanisms 52g Chemical Pattgrns 88-228. Cambridge Univ. Press. New York. NY. Ross. R.T. and Calvin. M. (1967). I; Eiophyg, 7. 595. Van Best. J.A. and Mathis. P. (1978). Bigghim, BiophzsI Acta. 503. 178-188. Trebst. A.. private communication. Jortner. J. (1980). Biochim. Biophys, ‘Agtg‘ 594. 193-230. Haan. G.A.. Duysens. L.N.N.. and Egberts. D.J.N. (1974). BioghimI BiOphysI 'Agtg‘ 8. 409-421. Auslander. I. and Junge. V. (1975). E§§§.L£££1 59- 310-315. RObinson. H.H.. Sharp. R.R.. Yocum. C.F. (1980). Biochgm, Biophyg, Res. CommunI 93. 755-761. Sun. A.S.K.. and Sauer. K.. ioghimI Bigphyg" 999;; (1971). 234. 503-508. Yerkes. C.T. (1979). M.S. Thesis. Th; Kinetigs g; 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 106 Electron Transport t3 Chlorgplast Photosystem ;; 41-44. Buttner. B. . Ph.D. Thesis Induced 52; Natural Delayed Luminesgenge gt Qrggn Plant Phototynthesis in press. 31-32. Govindjee. and Govindjee. R. (1975). Bioenergetics gt Photosygthgsis 1-150. ed.. Govindjee. R.. Academic Press. New York. NY. Pearlstein. R.M. (1982). Photosygthesitg ‘Vgt'; Engrgy Conversign ty Plants 32; Batteris 293-330. ed.. Govindjee. R.. Academic Press. New York. NY. Ke. B. (1972). Argh. Btothem. Eigphys, 152.70-77. NOrris. J.R.. Uphaus. R.A.. Crespi. H.. and Katz. 1.1.. (1971). Proc, BELLA; Acag, Sgi, ‘!§A 68. 625-628. Norris. J.R.. Scheer. H.. Druyan. M.E.. and Katz. 1.1.. (1974). Frog, Nat't. Agad. SciI ‘USA 71. 4897-4900. Davis. M.S.. Forman. A.. and Fajer. J. (1979). ProcI Ntt't. m Sci, 15576. 4170-4174. Iasielewski. M.R.. Norris. I.R.. Crespi. H.L.. and Harper. J. (1981). 1‘, Ag; .thgy ‘figgt 103. 7664-7665. Homan. 9.11.. (1973). Eur, L Bigghgm, 33. 247-252. Meites. L.. It; fienertt Multtptrtggtric gytytzfittttgg Prggrtg CFTfity (1976). Computing Laboratory. Dept. of Chemistry. Clarkson College of Technology. Potsdam. NY. publ. Malkin. R.. Beardon. J. (1975). EtothimI ‘fitgpgytt ARIEL 396. 250-259. Stiehl. 0.0.. witt. 11.1. (1969). _z_. Ntturfgrtgh, Tgil g 24. 1588. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 107 Haveman. 1.. Mathis. P.(1979). BtochimI Biophys, Acta. 440. 346-355. Malkin. R.. Bearden. A.J. (1975). Btothim Biophy ActaI 396. 250-259. Dutton. P.L. (1967). PhotochemI Photgtio. 24. 655-667. Babcock. G.T.. and Sauer. K. (1975). fiiochim, BiOphysll ActsI 375. 315-328. Joliet. P.. Barbieri. G.. and Chabaud. R. (1969). Photoghgml PhototigI 10. 309-329. Kok. B.. Forbush. B.. and McGloin. M. (1970). PhotgthemI Photgtio, 11. 457-475. Junge. M.. Renger. G.. and Auslander. I. (1979). FEBS Lttty 79. 155-159. Renger. a. (1977). anus‘tgtty 81. 223-228. Fowler. C.F. (1977). Eigchim, Bigphyt, Atttt 462. 414-421. Antonini. R.. and Brunoni. M.. gggggtgttg‘tngMyggtgttg tag Their Rttgtiont with Ligtnds (1971). Nerth-Holland publ.. Amsterdam. Blankenship. R.E.. Babcock. G.T.. and Sauer. K. (1975). Bigthim. Btophys, [Agtgt 387. 165-175. Blankenship. R.E.. Babcock. G.T.. and Sauer. K. (1976). BigthimI Btgphys, ‘Attty 396. 48-62. Babcock. G.T.. Blankenship. R.E.. and Sauer. K. (1976). FEBS Lgttt 61. 286-289. Renger. G.. and Weiss. I. (1976). FEES Lgttt 722. 1-11. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 108 Renger. G.. and Voelker. M. (1982). FEDS LettI 149. 203-207. Bong. Y.O.. Forster.V.. and Junge. I (1981).. FEES LettII 132. 247-251. Junge. M.. and 'itt. H.T. (1968). Z; Ntturfortch, 236. 244. Liptay. I. (1969). AgalI ChgmI 81. 195. Schliephake. M.. Junge. M.. and Mitt. R.T. (1968). gt Naturtorsch 236. 1571. Hong. Y.0.. and Junge. I. (1983). Bioghtg, Biophys. ActsI 722. 197-208. Theg. S.M.. and Junge. V. (1983). Bioghtg, Biophyt, Khanna. R.. Iagner. R.. Junge. R.. and Govindjee (1980). FEBS LottI 121. 222-224. Velthuys. 0.11. (1980). 1723 Lets.- 115. 167-170. Junge. M.. Auslander. M.. McGeer. A.J.. and Runge. T. (1979). Bioghim, 21221114, 33;; 546. 121-141. Johnson. J.D.. Pfister. V.R.. Hehmann. P.H. (1983). BiothimI BiOphys. ‘Attty 723. 256-265. Holten. D.. Hoganson. C..'indosr. M.R.. Schenck. C.C.. R.R..Migus. A.. Fork. R.L.. and Shank. C.V. (1980). BiochimI Biophys, .9299; 592. 461-477. 'ITT. R.T. (1979). Bigchim. ‘Etgphyty 'Attty 505. 355-427. Johnson. L.. HALQQQtMAQ c1979. public domain. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 109 Dye. 3.1... and Nicely. v.1. (1971). L. Chem, g1. 48. 433. Ghanotakis. D.F.. and Buttner. M.J. (unpublished data). Clayton. R.K. (1980). Phgtogygthesis: Phyttcal Mechanisms 929 Chemical Patterns 205. Cambridge Univ. Press. New York. NY. Reinwald. E.. Stiehl. H.H.. and Rumberg B. (1968). Z; Ntturtorsch 2365. 1616-1617. Amesz. J.. and De Grooth. B.G. (1976). BiothimI Biophyt, Attty 440. 301-313. Schmidt. 8.. Reich. R.. and Mitt. R.T. (1969). gt Naturforsch 246. 1428-1431. RS-232C INTERFACE FOR AIM 65. Document R6500 N08. Rev 1. July 1979. Rockwell Int'l Corp.. Newport Beach. CA. R6522 VERSATILE INTERFACE ADAPTER (VIA). Document 29000 D47. Order No D47. Rev 4. Nov. 1981. Rockwell Int'l Corp.. Newport Beach. CA. Buttner. B.. Ph.D. Thesis Ingucgt tgg Naturgl leayed Luminescengg gtmgtgtg Pttnt Phgtgtynthetis (in press). Dekker. J.P.. van Gorkom. H.J.. Brok. M.. and Ouwehand. L.. BigghimI Biophyt‘Agtgy (submitted). Joliet. P.. Delosme. R.. and Joliet. A. (1977). BiothtmI Biophyt, Act; 459. 47-57.