-43»- JE- - _ .'.“.T_‘9~.:-—- . . . J *1 h~< . . 3:44.», fix: ,g-r‘ng' ;4 $444.41 I p?“ - 4' "' ".\-".“IL'.‘4‘ ..?V‘..'.‘.I. 4: 44’4“ 1.4 ' .r;'5_5..J,§‘ 5 45"?" {.343} fi'dgl‘z" --",:rI-.‘ 4 ..'4"4 ‘ ’ A I > .' . ’ 414-14444", 444.4... 44.44.441.241... 5 ‘ '4 4 . . , .. :Q‘H 4.44.44. SEQ-444w, 4:214 4. 1 .-.. 5434“‘3'3..€o:.,;'j"?¢.-.-7{-215.5143444; "'I'. 5444454444424 ' L '2'“? “ In“. ‘T ' 4.4 444... 4115374 1-, 2:441 H.154}. 3| _4.44.’4‘14=..4I :IJ 4-14 , L -‘ a}. . ..- 4 ‘ “d74- .- 5. ' . . on. ‘v:;,j:I . . .4. . 5 .14 -.. - "- 4 ' ’~ rsfiiii'1345'fl‘f‘»4.143424444111455'4‘" " """~"*I‘4:"'444' $54443" 4'” '5‘?‘ '1‘" 04-4.: ‘1' 4 H{4J5 'Qfl .».,4-.~' .4- 175:4: '; 14' 44144421 ‘ ‘4 4‘4“. 4*. |.‘ 1", , 1‘ .444 $34.“ 0’ ' 44‘ “44""! 448444. ~94? ,' . .3'4 . .. JIX'} 40".”‘1‘7‘44 ‘4‘;— ' {.15 ',4. 4U "I: , 1'43‘I .I. AME: ag'ffib . .- :24 ‘ ‘ 9‘4! '0 \ln' J'P‘p-H “‘11—!" . . ' “43:34: ".4144 ' 434-4345 ‘34 4‘4" ”"4" '4' ‘24 ' ' '- -‘ I. .u a. ; 44V "‘~544J $449W4QMV.M4 W' kvflh " ' . r 4 u ' '13. o ‘ ‘. ' ._ 'l‘i' i“"5"r‘ "4‘4 "I:!'«‘.4';‘ (544 4- :4:4 “444:4? wj' 5 4 4 I 4.5.. '2" 4 r4 4444414563444“ IS: ._. 1. ‘ .o .u .5.- 1i- ‘4: - "Va .3 .. "$31335 .$ 4443‘? 4 :1 :4444. 4|.44Lm'“ 3“) {A}! "41"“ £4.61! ‘19" 4; e4 .4444 4 44¢ ‘44:}; 4,4..44‘4 4L .1421 5. 4;sz . 5 “'24 4343144 4 144...“. 441? . 1:4 4.. ‘ - ' ’hq 4.. 4 ; l .l. 4J4 [4'4 -. ‘ ‘4‘ £24 '4' 73"3‘F} 4; Ell-$443.1- 4644, ..;:. 7.;44‘ 44444‘4 ‘ ”42‘5”?” .5 v, - .4234'532 . ~ 52 {9'2“"3. ' ‘ :4 ".‘3' _ .- -')< . 74' 44 ‘ v.43 1.4 444.4:4' f4 .34.!» 4' .3 J"!- ’ :(r 1.? ,1’4‘>,I4 3v I 7 ..1-L| 1h ", |1' 14$ 4 4-14: .1 J41”; 1" 444 4 ' '414‘! 41"., ‘l. “‘4: 44-; I‘ .4 44: 4: I 4:4 41"" ‘t'114414)'4'444:‘4 ‘.2 'i L . 443.1. _ 144M... 44-;12‘ 4,... I 3.4-“. 34.444 (54.314: 4. ,N “ 1 '. i445§141|:. I("4'14 404' ' ' n4" .§ ,4. .4.“ 4 . '1 .144' !"il'..‘ .44434‘444'; LI'm“. .. (4 m4 . 4 I ; 4‘. .. J4. *- - .414 . , a w::.j{‘.f-44‘Zf4"1 ~ . 2'4 4T.'41'03 1.’ = 125. There are four transitions with total angular momentum L; = :I:l and 21:9; each transition gives rise to one state with S = 0 and three states with S = 1. The corresponding 16 excited states will have energy transitions of (h2/41tmR2)(52 - 42). A ring of radius 4 A gives a transition energy congruent with 580 nm light. Thus, from Sirnpsons theoretical model a ring the size of a porphyrin is predicted to have transitions in the region that they are actually seen in experimentally. From Hund's rule we can qualitatively predict that that the lowest excited energy state should be a triplet, and that the singlet states with L; = :I:1 should have higher energy than those with L; = i9. Since the ground state is a closed shell with L; = i0, the normal selection ALz = :t1 predicts that the higher energy singlet is allowed and the lower energy singlet is forbidden. With these arguments the L; = :I:l singlet excited state was assigned to the Soret band and the L; = i9 singlet excited state was assigned to the visible bands. In 1954, Moffitt extended Simpson's studies to a cyclic polyene and obtained the same results as Simpson with the added benefits of correctly explaining the Qx and Qy energy levels [2c]. In 1950, Longuet-Higgins, Rector and Platt used MO calculations and reported the am and a2u orbitals, in D411 notation, as the HOMOs and the cg orbitals as the LUMOs [2d]. These orbitals are analagous to the free electron orbitals l = i4 and l = :I:5 [2e, 10a]. An - electron, excited to a higher energy level, can be either an a2u —> eg and am —> eg transition. The excited state configugurations are denoted (azueg) and (a2ueg). These assignments were consistent with the corresponding wavelengths seen in the optical spectra of porphyrins. However, as a shortcoming the theoretical models still, incorrectly, predicted equal intensities for both the Q and B bands. In 1959, Gouterman added configuration interaction (CI) to the pioneering work of Simpson's free electron model, Moffitt's cyclic polyene and Longuet-Higgins and coworkers' molecular orbital (MO) calculations. This has came to be known as Gouterrnan's four orbital model. 1.2.1.1 Four Orbital Model The four orbital model is based on the simple HOckel model developed by Simpson and Moffitt, combined with the MO calculations of Longuet-Higgins. This model considers the two nearly degenerate HOMO orbitals labeled by Gouterman as b1 and b2 and are of a2u and am symmetries, respectively. The use of the b1 and b2 labels are due to the fact that the assigned a2u and a1“ symmetries are only appropriate for the full D41. symmetric porphyrin. However, 10 Figure 1-4 Pictoral representation of the HOMOs (b1 and b2) and LUMOs (c1 and c2) of the porphyrin four essential orbitals. The size of the circles are proportional to the orbital coefficients. Solid circles represent positive values and dashed circles represent negative values. The heavy lines indicate nodal planes. ' Figure 1—5 Y i ii PORPHINE “"3 “—co--— —-—- I I LJ ‘ OPP-(THP) CHLORIN cl - ~-. h—J b, aDJ-(THP) Relative orbital energy representations for porphyrin and chlorin. An equivalent energy d has been added to bl and subtracted from b2 to make these orbitals degenerate in the porphyrins. 12 as mentioned above, the four orbital model has been taken as applicable to all four pyrrole macrocycles. The degenerate LUMO orbitals labeled by Gouterman as C] and c2 are each of eg symmetry in the full D41, symmetric porphyrin. The c1 and c2 labels are applicable to pyrroles that are not of D411 symmetry. The orbitals are represented schematically in Figure 4. The atomic orbital (A0) coefficients are proportional to the size of the circles. The dashed circles show a change in sign and the heavy lines represent nodal planes. The approach Gouterman took to combine the partial explanations of the previous workers was to assume CI as of prime importance to explain the optical properties of the porphyrins. The other assumptions in Gouterman's considerations were that full D411 symmetry is in effect and that the HOMO orbitals are accidently degenerate. It was shown that these assumptions were valid for explaining the porphyrin and porphyrinoid structures. If the bl and b2 orbitals are degenerate then electron interaction would cause the singlet transitions to mix as follows: 52—} =1th —> c.) i (b. —> calm-t 3; } _ -i Q" — [(b, -* c,):I: (b2 -) 02)](2) I Figure 5a is a representative diagram of the nearly degenerate case. Figure 5b is a diagram of the transitions if the x and y degeneracy is lifted. In the case of chlorins, which has no x,y degeneracy, we might expect that the x-polarized states would still be defined by eq. 1 because the two transitions remain nearly degenerate. However, the y-polarized states would become: 13 B, = cos 11(191 -> c1) + sin 11(b2 —9 c2) Q, = sin 17(bt -> ct) - cos 71(1): -> cs) where n is choosen so that the lower energy transition (b2—>c2) is more heavily weighted in the lower energy state Qy. If (bl—>c1) and (b2—>cz) have roughly equal transition dipoles, then the ratio of absorption strengths will be: £’(Q,)=1-sin2n (2(a) 1+sin2n eq.3 In eq. 3 we see that as the energy of the two tansitions (b1—>c1) and (b2—>cz) become increasingly unequal, 11 drops below n/4, and the Q: band gains intensity at the expense of the By band. The orbital energies of Fig. 5 are essentially those of a simple Hiickel calculation with a 5 energy added to b2 and subtracted from b1 to make b2 and b1 fully degenerate. Since the two x-polarized transitions remain nearly degenerate the wavelengths and absorption strengths of Qx and Bx remain similar even when the x and y degeneracy is no longer present. However, as the x and y degeneracy is lost the Qy band will red shift and become intensified. This is very evident in the optical spectra of chlorins. Thus, it is seen that the four orbital model can be used to explain the observed optical spectra of ' porphyrins and porphyrinoid compounds. 1.2.2 Spectroscopic Significance Electronic absorption spectroscopy of porphyrins can be explained by the four orbital model. Electronic transitions, are produced by the absorbance of light and give either B or Q bands. The B band, also known as the Soret band, comes from the allowed transition of the 1: to 1t* state. The Q bands in the visible region 14 come from configuration interaction of the four orbitals with some vibronic ~ coupling of the Q bands with the Soret band [12]. In Q bands there are transitions of the Q(0,0) referred to as the a band. The symbolism Q(0,0), used here, represents the electronic transition of the ground state at the lowest vibrational level to the excited state lowest vibrational state. To the blue side of the (1 bands are transitions of Q(O,l) known as the [3 bands. The Q(0,1) symbol would represent the electronic transition of the ground state v = 0 to the excited state v = l and so on. In a few highly coupled porphyrins the Q(0,2) transitions are also discernible. In the Soret band the transitions are less obvious owing to the intensity of the B(0,0) band. The B(0,l) is seen as a shoulder on the blue side of the Soret band. The visible bands, on the other hand, are readily resolvable into the various transitions. These transitions, although weakly allowed, have approximately one tenth the intensity of the Soret band. In addition to the visible bands of the porphyrin, metalloporphyrins have charge transfer bands that are, as a rule, not observable. Notable cases of observable charge transfer bands are found in Mn3+ etioporphyrin I [8a, 8b], high spin Fe3+ metalloporphyrins [8c] and most importantly in cytotochrome c oxidase spectra [9]. Because all of the optical absorption bands fall in the visible light region we can complement optical spectroscopy with the use of Raman spectroscopy. Raman spectroscopy has become particularly useful in elucudating protein structure and functionality[7b, 13]. When Raman spectra are taken so that the excitaiton wavelength is within one of the optical absorption bands, we can obtain Raman spectra that have enhanced vibrational modes related to the specific optical band. In other words, we can isolate chromophoric properties related to specific optical transitions. 15 1.2.3 Raman Theory An explanation of Raman spectroscopy has its origin in classical theory [13]. If we start with a light wave of frequency 1) and an associated electric field of strength E that is dependent on frequency ‘0, then: B = AcosZmrt eq. 4 where A represents the amplitude and t represents the time. If we use the light associated with eq. 4 to irradiate a diatomic molecule, we can calculate the induced dipole moment from: P = (E = aAcosZmrt eq. 5 where or represents the polarizability. A molecule vibrating at frequency 01 has a nuclear displacement q that can be calculated from: q = qocos21tu1t eq. 6 where qo represents the vibrational amplitude. If vibrational amplitudes are small, then a is a linear function of q. Thus: a = are + q(3a/3q)o eq- 7 where (10 represents the polarizability at the equilibrium position and (Bot/8mg is the rate of change in polarizability with respect to the change in vibrational 16 amplitude, evaluated at the equilibrium position. We can now combine equations 1-3 and obtain the following: P = aAocos21mt .-. aOAOCOSth + (Ba/Bq)oquocos21tut cos21m1t = avocos21mt + 1/2(80/8q)oquo{cos[21t(o + out] + cos[21t('u - 001]} eq. 8 From classical theory we can assign the terms of eq. 8 as follows. The first term represents an oscillating dipole that radiates light at frequency 1); this is known as Rayleigh scattering. The second term describes the Raman scattering of light frequencies 1) - ‘01 and u + tn . These terms are known as Stokes and anti-Stokes scattering, respectively. From eq. 8, we see that if (Ba/8q)o goes to zero then the second term goes to zero and there is no Raman scattering. In other words, if the polarizibility of a molecule does not change during the vibration, then the molecule is not Raman active. Figure 6 is a diagram representing some of the possible effects of a photon-molecule interaction. The Stokes lines arise when an excitaion light source of u is impinged on a molecule in the v = 0, or ground vibrational state. The frequency of the scattered light (us) that is collected by the detector is equal to the frequency of the exciting laser line (1)) minus the frequency (1)1) cooresponding to the molecular transition of the ground vibrational state to first vibrational state of v = l. The anti-Stokes line is the frequency of the exciting laser line plus the frequency cooresponding to the energy difference between the first vibrational state and the ground vibrational state. It is in fact, true that the anti-Stokes lines have the same vibrational information as the Stokes lines, however at a much lower intensity. The lowered intensity of the anti-Stokes lines can be explained by the Maxwell-Boltzman l7 - \ ° : 4 l W 3 Upper (excited) 2 etectromc state I t O 1 VI ’ anti-Stokes E'I Stokes " o ‘5 I '0 IO. U“. '0‘ v“. V" U” 8r 7*. '3‘ V“ I It“ 11 I) 4 II 11 y 41 _ - U _ U“ 5 U. N U: D! '3 z 2 i 1 Lower (amend) ‘ r electronic state 1 g, t. Infrared Rayleigh Narmol Prerescnonce Resonance Finerescence v” J Roman Roman Roman Figure 16 Some possible consequences of a photon of light interacting with a molecule. The lengths of the arrows are proportional to the frequency of light that is either incoming (upward), scattered (downward) or emitted (downward).(Reference 7c). 18 distribution law. That is, the number of molecules in the 'v = 0 vibrational state by far exceedes the number in the v = 1, or first vibrational state. Due to this fact all the Raman spectra shown in this thesis are Stokes scattering Raman spectra. 1.2.3.1 Resonance Raman The theory given above for Raman theory also applies to resonance Rarnan spectroscopy (RR). The major difference between the two types of spectroscopy is that in standard Raman the energy of the excitation line falls below the enery level nescessary to excite the molecule into the the excited state. However, in R the frequency of the excitation laser line is of a wavelength that is of sufficient energy to place the molecule into the excited state. Although conceptually this could be mistaken for resonance fluorescence (RF), there are some major experimental differences. One in particular applies to RR of porphyrins and that is that RF lines are all polarized, and in RR spectroscopy some lines are polarized and some lines are depolarized. The classical theory of Raman spectroscopy is sufficient to explain the energy of the scattered light, however, it does not explain the Raman vibrational band intensity [7c]. To explain the intensity I, we must incorporate quantum theory. For the Stokes process we will start with the classical expression for the Raman intensity of the scattered light: I=K(Uo- v».-t)‘lo(datuldr)2 eq. 9 where K is a constant, 10 is the incident light intensity, and r is the distance separating nuclei m1 and m2. From eq. 9 we see that the intensity of the Raman light is proportional to the fouth power of the scattered light. It is also apparent from eq. 6 that the intensity is dependent on the change in polarizability of the 19 molecule. The change in polarizability is the basis of the bond polarizabity theory [7d], which tries to explain Raman intensity from the classical approach. However, the theory is concerned only with the properties of the ground electronic state. This theory works well for Raman intensities that are far from resonance, but inappropriate for eXplaining resonance Raman intensity. In the quantum mechanical treatment we have the following: I = K(v. :I: M)‘IoZKam)~.|2 eq. 10 where omn represents the frequency resulting from a molecular transition from state m to n and laqplmn is the transition polarizability tensor . (OEOP)"' = i2 ("WWW“) + finludlrxrlwlm) , the—trout vm+vo+rTr eq.11 In equations 10 and 11 the molecule is considered, initially, to be in the molecular state m. An electromagnetic wave of frequency 00 and intensity 10 causes the transition into the n state, while scattering light of frequency 00 i um“. The sum over r includes all of the quantum mechanical quantum eigenstates of the molecule, h is Plank's constant and I“; is a damping constant which takes into account the finite lifetime of each molecular state. The amplitudes of the electric dipole transition moments is represented by the integrals , etc.; where lip is the electric dipole moment operator along direction p. In eq. 11 we see that as no approaches the energy of an allowed molecular transition um, (um) - pa + iI‘r) becomes small; consequently, one term in the sum becomes very large. This is, in fact, the resonance condition. If we consider the Born-Oppenheimer 20 approximation, we can write the molecular states m and n as the product of their electronic and vibrational states. Thus, the total wavefunctions can written as: lm> = lg)|i>, In) = lg)|j>, Ir) = le)lv> eq. 12 where, g is the ground state and e is the electronic excited state. The i and j terms represent vibrational states associated with g and v is a vibrational state associated with e. If we now refer to Fig. 6 we can see a correlation with the levels depicted. The lower electronic state corresponds to g and the upper electronic state corresponds to e. The i and j terms are v" levels in g and v corresponds to a v' level in e. If we use eq. 12 we can write the dipole transition moments as: = (leelw and (nnolm = eq. 13 where Me, which is the pure electronic transition moment from the ground to the eth excited electronic state. Now, the expressions on the right hand side of the equal signs in equation 10 involve only vibrational wave functions. The electronic wave functions appear in Me. and Me changes as the nuclear coordinates vary. We can account for these changes by expanding the electronic transition dipole moment in a Taylor series about the equilibrium geometry as a function of the normal coordinate Q. Hence, we have: Me=Me+(3Me/3Q)°Q+“° 94-14 21 There are equivalent expressions for each of the 3N-6 possible normal A coordinates. If we now substitute eq. 14 into eq. 13 and subsequently substitute eq. 13 into eq. 11, we get: (aap)m=A+B eq. 15 where MflMl (jlevIr') =_L__L A h Zvu-w-i-il‘v ”'16 and fl 2 ° - - B: M e (6M c laQ) ZQIVXVIQII) eq.17 h , w-v.+rT. The term ovi represents the energy gap between the ith vibrational level in the ground state and vth vibrational level in the excited resonant electronic state. In the right hand side of eq. 16 we have ignored the nonresonant term and the term for which p and o are reversed has been omitted. ' The dominant contributor to the resonance Raman intensity is the A term. This is the case since, for allowed transitions Me, the electronic transition moment, is larger than the change in the dipole moment with respect to the normal coorrdinate Q (aMclaQ). The vibrational overlap integrals, and (also known as the Franck-Condon factors) are equal to zero for non-totally-symmetric modes, therefore, only the totally symmetric modes give rise to resonance enhancement. The A term involves a single electronic state; however, the B term arises from the vibronic mixing of two excited states. The B term active vibration may have any symmetry that is contained in the direct product of the group- theoretical representations of the two electronic states. The vibronic mixing is an 22 important feature of the B term, however, the major importance of the B term mechanism is the intensity it allows to scattering by non-totally-symmetric modes. What this means is that normally weak vibrations can be observed because they are able to borrow intensity from nearby intense transitions. We can now make some generalizations about the use of Raman spectroscopy. Equation 9 defines the classical theory of scattered light intensity. We can see from eq. 9 that Raman intensity, far from resonance, is mostly dependent on polarizability of the molecule. That is, the Raman intensity depends on the degree of change of the polarizability caused by a change in the normal coordinate. In resonance Raman there are two terms that add intensity to ' the Raman scattering. The A term mechanism involves a single excited electronic state, whereas the B term mechanism involves two excited electronic states. The combined results of these terms is that resonance Raman intensity may be enhanced as much as 104 over the normal Raman intensity. The large intensity enhancements are expected if, for both terms, the excitation frequency is close to the electronic transition, and the absorption intensity of the electronic transition is large. Both terms also experience enhancement if the geometric changes experienced during a particular normal mode are similar to those changes occurring when going from the ground to the excited electronic states. For the A term only some of the products of the Franck-Condon factors and are large and for the B term, only, the vibration is effective in mixing the two excited electronic states. The intensity gained by resonance Raman spectroscopy over normal Raman spectroscopy, makes it an ideally suited spectroscopic technique to study porphyrins in both model and biological systems [7b, 13c]. The principle of resonance Raman gives us the opportunity to enhance vibrational properties relevant to a specific electronic absorption band. For example, ligands may be 23 vibronically coupled to the porphyrin 1t system. So, excitation into the Soret band allows detectability of the ligand by its vibrational frequency and corresponding isotope shift. If, however, the ligand is not vibrationally coupled to the porphyrin 1t system, then excitation into the Soret band would not enhance the ligand vibrations. However, excitation into one of the Q bands or a charge transfer band could well enhance vibrations of the ligand. This technique allows us to probe a mixture of compounds and isolate vibrational modes belonging to one specific compound; or in the case of a protein to isolate the active site that contains the porphyrin structure. 1.2.3.2 Established Porphyrin Vibrations Figure 2 shows the label designations for the carbons of the porphyrins. The carbons directly bonded to the nitrogens are known as the alpha carbons (Cat). The backbone carbons of the pyrrole rings are designated the beta carbons (C5) and the bridging carbons are referred to as the meso carbons (Cm). The carbon-nitrogen skeleton that makes up the basic porphyrin has been extensively studied. The various vibrational modes are well known and their sensitivities, although not fully understood, are well characterized. The RR bands in the 1000 cm'1 to 1700 cm‘1 region correspond to the in-plane stretching of C-C and C-N bonds. The bands above 1000 cm“1 are not only very sensitive to the peripheral substituents but also indicative of a number of effects on the central metal or macrocycle moiety. Normal coordinate analyses of nickel [14a], copper [14b], and iron [14c] porphyrins have been used to establish both the RR and IR active vibrational modes of a number of porphyrins. The calculated frequencies, combined with specific isotopic substitution, has allowed for frequency and symmetry assignments of these porphyrins. The normal mode calculations and RR work of 24 Abe et al.[14a] established the frequency labels that are used in this thesis. The other important information from the work of Abe et a1. is that the Cp-ethyl vibrations only contribute to the porphyrin modes. In other words the internal vibrations of the ethyl groups do not affect the porphyrin vibrations. From an experimental standing this means that substitution of the ethyl groups with other alkyl groups has little or no effect on the porphyrin vibrations. Typical porphyrin vibrations are generally a mixture of vibrational motions. This is best seen in Figure 7, which shows the basic vibrational motions of the individual atoms for a given frequency assignment. With the mixing of vibrational modes we typically get several vibrational bands that are each sensitive to metalloporphyrin properties. These coupled vibrational bands follow certain systematic trends that give us an opportunity to characterize a variety of metalloporphyrin traits. Some of the more informative characterizations that we can get from RR spectra are the distinction between five and six coordinated porphyrin complexes, the oxidation and spin states of the central metal and many environmental factors that affect the coordinated ligands and/or the metal or the porphyrin itself. A Specific Raman active vibrations of interest are the vibrational modes known as 02, D3, D4, 010, ‘01] and 1919. The designated vibrational modes come from the assignments of Abe et alii. These modes are known as the oxidation state marker ( u4), core size markers ( 03, D10 and U19) and spin state markers (02 and 011). The major contributing factors associated with these vibrations are 0(CaN), 0(CaCm) and MCBCB), respectively. 25 Figure 1-7 Porphyrin vibrational modes that are sensitive to the oxidation or spin state of the central metal or the core size of the the porphyrin itself. vibrational values given are for NiOEP in methylene chloride. (Reference 13). v7 (A19) 674* K I! “a 5 fig l v11(819) 1576 V4 (A19) 1383 figs/RV V . N >/ '\ 5:1”; vmtelgt 1655 V19( A29) 1603 27 The oxidation state marker band ‘04 is the most intense porphyrin vibration observed when the porphyrin is excited near or in the Soret band. It is named, accordingly, due to it's sensitivity to the oxidation state of the central metel. The U4 vibration is particularly sensitive to the oxidation state of iron porphyrins. In ferric heme proteins 04 is detected at approximately 1375 cm'l, while in ferrous heme proteins 04 is located at approximatly 1355 cm'l. All the vibrational modes above 1450 cm'1 have an inverse linear relationship with core size. The modes that are particularly sensitive to the core size are the ones that have their major potential energy distribution in the Cor-Cm vibrations. The core size markers and particularly ‘010 are good indicator bands for distinguishing between five-or six-coordinated metalloporphyrins. This is, in effect, due to the out-of-plane nature of the five coordinate complexes. In five coordinated complexes the ligand pulls the metal out of the porphyrin skeletal plane. The metal nitrogen bonds are subsequently drawn toward the metal, which causes the pyrrole ring to tilt. This does, in one sense, increase the core size of the porphyrin which does of course affect the core size marker bands. However, tilting of the pyrrole rings also causes a torsional strain on the meso carbon and a carbon bonds that is detectable in the vibrational modes that have their major potential energy distribution contributed by these bonds. 1.3 Core Metals and Oxidation States As mentioned earlier, the functionality of iron ranges from 02 transportation in hemoglobin and O2 storage in myoglobin to electron transportation in the oxidase cycle of cytochrome c oxidase (CcO). During the reduction of 02 to H20 in CcO heme a3, coupled in proximity with CuB, receives an electron from cyt. c, through heme a and CuA to become an Fen(Figure 8). Figure 1-8 Simplified mechanism for the reduction of O2 and proton translocation in cytochrome oxidase. The binuclear center is represented as [a3 CuB]. The reaction is initiated by 02 binding to the reduced form of this center, to produce oxy, and then peroxy intermediates. Reduction and protonation of the peroxy species is coupled to the translocation of two protons, which is indicated by the heavy, shaded arrow. Oxygen-oxygen bond cleavage is also driven, in this step, to produce the ferryl intermediate. Reduction and protonation of this species is also coupled to the translocation of two protons, which is indicated by the second heavy, shaded arrow. The hydroxy product of this reaction is subsequently reduced by two electrons, to complete the cycle.(Figure courtesy of G. T. Babcock). 30 2 Cyt-C (Fem) r 2 Cyt-C are“; Figure 1-9 Proposed mechanism for the formation of compound 1 of Cytochrom c peroxidase. a) Representation of the native enzyme. b) The aetivated complex with the distal histidine acting as an acid-base catalyst and the active site argenine stabilizing a partial negative charge forming on RO-OFe. c) The oxoferryl n-cation radical intermediate. (1) Compound I after the intramolecular electron transfer to to produce the oxoferryl and free radical X.(Reference 19). 31 Subsequent steps of electron and proton transfer take heme a3 to an 0=FeIV complex [18]. Peroxidase binds peroxide (H202) in the ferric form. Subsequent oxidation produces an oxoferryl cation radical that is the active oxidizing form of the enzyme that carries out nonspecific oxidation of a number of substrates (Figure 9)[19]. Catalases react the same as peroxidases in their initial steps. They are never found in the ferrous state naturally and virtually impossible to reduce to the ferrous form. The major difference between the catalayses and peroxidases is that the catalayses are very substrate specific. Cytochrome P450 is a drug metabolizing defense mechanism in animal life. Although not all of the intermediates are fully characterized it is reasonably accepted that the highly oxidizing intermediate is also an O=FeIVP+- complex [19]. Virtually every metal has, at some time, been inserted into synthetic porphyrins. The usefulness of metalloporphyrins not naturally found in biology are many fold. For example, oxoferryl model compounds are unstable above -30°C, [45] so oxovanadyl porphyrins have been used to study environmental effects on the oxo ligand [46]. Zinc and magnesium octaethyl and tetraphenyl porphyrins, for example, were instrumental in setting the ground work in the study of n-cation radicals. 1.3.1 n-Cation Radicals One electron oxidation of metalloporphyrins may take place at either the core metal or the n—system of the macrocycle. The major factors that influence the site from which the electron is removed are the substitution pattern on the macrocycle skeleton, the central metal, and environmental conditions. Much attention has been given to the understanding, spectroscopic characterization, and predictability of which orbital the electron is removed from when a n-cation radical is formed. In the formation of a 1t-cation radical, it is possible to remove an 32 electron from either one of the HOMO's (all. and 32“), depending on which is lower in energy for a given MP. Examples of 1t-cation radicals of the 2A1“ and 2A2“ forms are both known from the work of Dolphin and coworkers [15a]. From ESR studies it was possible, to first, establish the oxidized products of zinc and magnesium octaethyl and tetraphenyl porphyrins as n-cation radicals. From the electron density distribution of the ESR work it was also possible to determine which of the n-cation radical types were formed. It was established that both ZnOEP and MgOEP had 2A1“ ground electron states, while ZnTPP and MgTPP both had 2A2“ ground electron states. The established ground electron states have unique optical spectroscopic properties. The optical properties and subsequent optical spectra are referred to as A11, for the 2A1“ ground electron state and A211 for the 2A21. ground electron state. The work of Dolphin and coworkers was intrumental in establishing the formation of 1t-cation radicals in model compounds and then relating the model compounds to optical spectroscopic patterns in biological intermediates [15b]. Many types of spectroscopic techniques that have been used to characterize rt—cation radicals. In addition to the optical and ESR techniques mentioned above, there are MCD [29], NMR [30], IR [31], X—ray crystallography [32] and RR, which will be focused on here. Detailed RR studies of n-cation radicals have identified the structually sensitive vibrational modes [16, 11a]. The vibrational modes sensitive to oxidation of the macrocycle, as one might expect, are the same modes that are structurally informative in the parent porphyrin. These modes are 02, D3, 04, 010, ,011, and 1319; their vibrational character was described earlier. The work of Oertling et a1. [11a] on MOEP+- (M = Cu, Zn, Co) [11] and the work of Czemuszewicz et. al. on MTPPi- [16b] and MOEP+- (M = Cu, Ni and Fe) has shown that the the structually sensitive Raman bands of these porphyrin 33 cation radicals have systematic trends. For the octaethyporphyrins the modes involving primarily "CaCm stretching character (1)3 and U10) decreased in frequency relative to the parent compound. Those bands whose primary stretching character was in the CBCB portion of the ring (U11 and 02) increased in frequency relative to the parent compound. The frequency of D4, which is primarily a CaN stretch, decreased relative to the parent. The vibrational modes for the cation radicals of MTPP shift somewhat differently than the corresponding modes of the cation radicals of MOEP. Particularly, the mode 02, that increased in the MOEP+-, decreased in the MTPP+- case. The down shift of '04 is much less in the MTPP+- compared to the down shift of the MOEP+-. These differences in mode shifts has been explained in terms of the an, and azu orbitals being antibonding and bonding, respectively, with respect to the CpCB bonds. 1.4 Environmental factors A number of factors influence the binding properties a meta110porphyrins in model studies. The effects on the model compounds can and often do give invaluable insights into the protein’s effect on the macrocycle active site. Among some of the factors that effect model compound bindings are temperature, solvent properties and atmosphere. 1.4.1 Solvent When metalloporphyrins are oxidized, solvent effects are a key consideration of where the oxidation takes place [22]. For example, oxidation of metalloporphyrins in coordinating solvents causes oxidation to occur at the metal center. The solvent often becomes either a mono or in most cases, a bis coordinated metalloporphyrin (LanPXn :z 1 or 2) complex. Oxidation of 34 metalloporphyrins in noncoordinating solvents can, and often does produce metalloporphyrin n-cation radicals (MP+-). When metalloporphyrin n-cation radicals are subsequently exposed to coordinating solvents, or other potentially coordinating media, they can undergo an intramolecular electron transfer to give an (LanPXn = 1 or 2) complex. Solvent binding to metalloporphyrins can give an understanding of the competitive and hindering effects of water soluble ligands that may be exposed to the active site of a protein. For the most part, aqueous ligands are anions and typically inhibitors of CcO. One of the most commonly studied and best known of these aqueous ions is cyanide. Cyanide (CN') binds to the FeIII state much more rapidly and stronger than to the Fe11 state of hemoproteins. This is easily understood due to the fact that a ferric iron state in a porphyrin must be in close proximity, if not actually bound, to an anion. The cyanide anion is a relatively strong anion in terms of nucleophilicity. It, subseqently, is tightly bound to a ferric state iron porphyrin and only weakly attracted and bound to a ferrous iron porphyrin. Once bound to the ferric state of an oxidase protein removal is virtually impossible since the ferrous state of a hemoglobin is the only exchange protein available to oxidase [4]. This results in cyanide being highly toxic to any form of animal life. 1.4.2 Atmosphere A number of metalloporphyrins, when oxidized under an inert atmosphere such as argon or nitrogen, produce a metalloporphyrin n-cation radical (MP+-). If, however, they are oxidized under an atmosphere that has coordinating capabilities such as carbon monoxide (CO), nitrous oxide (NO), oxygen (02), etc., then ligated metal oxidized MP's are formed[23]. Understanding the binding property of gaseous ligands is essential since biological proteins are subjected to 35 these ligands. Gaseous ligands are in constant competition with 02 for binding to the metal core of active site hemoproteins. In hemoglobin, for example, carbon monoxide is known to have a much stronger bonding affinity than oxygen and yet the protein moity functions to discriminate for 02 binding over CO binding [4, 5]. Debate over the protein's ability to bring about this discrimination has given rise to a variety of model compounds. These models can affect the binding site by either a strap [24], a cap [25], a picket fence [26], even cofacial or face to face diporphyrins [27]. Therefore, the search for a "true" model system and definitive experiment are ever important. 1.4.3 Temperature It might, at first thought, seem that temperature effects on the binding properties of metalloporphyrins has little relationship to biology. For the most part biological systems carry out chemical processes at a controlled and constant temperature. However, since it is virtually impossible to perform isolated ligand binding and spectroscopic studies in vivo the effects of temperature on metalloporphyrin binding rates and strengths can produce invaluable results in understanding the role of the protein that surrounds the active site macrocycle. Temperature is a major factor in the ligating properties of metalloporphyrin model compounds. Exposure of reduced metalloporphyrins to oxygen at high temperatures (typically above 0°C) results in oxidation of the metal. If, however, reduced metalloporphyrins are exposed to oxygen at low temperatures they can form a semi-stable 02 bound metalloporphyrin complex. The binding abilities of oxidized MP's are sometimes temperature dependent. For example, H20 binds reversibly to CoIIOEP+-(ClO4’) only at low temperatures. Methanol, however, readily binds to the CoIIOEP+-(C104)‘ at room temperature to give a[(MeOH)2ComOEP]+(ClO4)' complex[21]. 36 1.4.4 Hydrogen bonding One of the most important intermolecular forces in chemistry and biology is hydrogen bonding [33]. It is well known that hydrogen bonding is one of the major influences on protein folding [34]. The influence of hydrogen bonding at the active sites of hemoproteins is, however, an issue of much debate [35]. Hemoglobin and myoglobin are subject to binding by ligands that they are in constant contact with. The question that has yet to be answered is what mechanism the protein uses to actively discriminate between oxygen and other available ligands that are better at liganding the bare heme [36]. In cytochrome c oxidase the possibility of a hydrogen bonded oxoferryl intermediate has sparked controversy and is at this time an unresolved issue [37-38]. The true nature of the hydrogen bond is in itself an unresolved issue. The classical definition of the hydrogen bond, given by Pauling described it as ionic/electrostsatic in nature [33]. At one time it was believed that only complexes consisting of the most electronegative elements nitrogen, oxygen or fluorine could be be involved in hydrogen bonding. The ongoing work of Green et a1. [45] and Zwier et al. [39] has shown that not only strong electronegative atoms could be involved in hydrogen bonding, but also that 1t-electron clouds of cyclic polyenes could be hydrogen bond acceptors. These workers not only shed new light into the nature of the hydrogen bond but often relate their findings to biological systems with convincing results [40]. The work of Zwier has shown that chlorohydrocarbons could even be hydrogen bond donors [39a]. With such an active interest in hydrogen bonding in protein structure control and the previously reported cases of potential hydrogen bonding at active site ligation it is only conceivable that this work be carried into the field of biological model compounds. 37 Intermolecular hydrogen bonding studies have been done for a number of compounds under a variety of conditions. Cobalt substituted myoglobin or hemoglobin bound with oxygen shows a deuterium isotope effect when H20 is substituted with D20 [41]. Carbon monoxide bound to ferrous cytochrome c is also known to exhibit a deuterium isotope effect on ‘D(CO) when D20 is used relative to H2O [42-43]. The studies mentioned above were all done with either homoglobin or myoglobin. However, the possibility and possible role of hydrogen bonding being an active participant in reaction intermediates of ‘ oxidases has yet to be answered. Chang and co-workers have presented a series of compounds where one of the meso carbons is substituted with either a naphthyl or anthryl unit that may contains a functional group capable of effecting intramolecular hydrogen bonding with ligands coordinated to the core metal. Alterations of the functional group from acid to amide to alcohol to ester have given orders of magnitude decreases in 02 binding ability as measured by titrative methods. This model, however, has a free rotation bond that does not allow for consistant hydrogen bonding effects or lack there of. To lock the functional group into a position that is consistantly in the vicinity of a ligand binding to the metalloporphyrin pocket site they have created the Kemp's acid structure (Figure 10c). This system, although, not absolutely rigid does allow for only minimal displacment of the acid, amide, alcohol or ester group. The titration results, again show orders of magnitude increases in 02 binding over any previous 02 binding models. 38 Figure 1-10 (a) Structures of OEP and TPP, synthetic porphyrins that have been well established as models for a variety of biologically related structures. (b) The structure of the recentlly synthesized naphthyl Kemp's acid etio type porphyrin (NKAP) model compound that is capable of intramolecular hydrogen bonding. \ C) 40 1.5 Aims Of This Research My research has been oriented toward the use of model compounds to gain an understanding of the role of porphyrin compounds in biological systems. Although some projects used chlorins [47], biological samples [49], or involved the synthesis of non-porphyrin compounds. The theme of this dissertation will focus on the porphyrin chemisu'y upon which I concentrated. The parent [11a] structures that are the focus of the majority of the work described in this thesis are octaethylporphyrin (OEP), tetraphenylporphyrin (TPP) and naphthyl Kemp's acid porphyrin (NKAP) (Figure 10). Synthesis of NKAP and it's derivatives was done by Ying Liang and details of the process will be described in her dissertation and/or corresponding publications. The core metals included cobalt, iron, vanadium, ruthenium and aluminum. The binding ligands were the following: carbon monoxide(CO), oxygen (02), oxo(O=) and cyanide(’CN). Specific compounds used in my studies are given detailed explanations within the following chapters of this dissertation. Specifically, Chapter 2 will concentrate on n-cation radicals. Chapter 3 will be dedicated to hydrogen bonding. Chapter 4 will be a study in cyanide binding to fluoro-substituted phenyls of tetraphenyl iron porphyrins and Chapter 5 will be a summary of additional projects. Projects that will not be discussed in this dissertation, although published or submitted, included chlorins,[47] biological samples,[49] and transient lifetimes of manganese OEP cation radical[48]. 41 REFERENCES [1] (a) Dolphin,D. (Ed.); The Porphyrins. Academic: New York Vols. VI and VII [2] (a) Gouterman, M. J. Chem. Phys. 1959, 30, 1139-1161 (b)Simpson, W.T. J. Chem. Phys. 1949, 17, 1218- 1221 (c) Moffitt, W. J. Chem. Phys. 1954, 22, 1820- 1829 (d) Longuet-Higgins, Rector, and Platt J. Chem. Phys. 1950, 18, 1174-1181 (e) Platt, J.R. J. Chem. Phys., 1954, 22, 1448-1455 [3] Metal ions in Biological Systems; Studies of some Biochemical and Environmental Problems, Plenum Press, New York 1973 [14] Stryer, L Biochemistry, 3rd Ed., W. H. Freeman and Company, New York 988 [5] Springer, B.A., Sligar, S.G., Olson, LS. and Phillips, G.N Jr. Chem. Rev., 1994, 94, 699-714 [6] (a) Pratt, J.M. Inorganic Chemistry of Vitamin BIZ, Academic Press, London, 1972 (b) Herlinger, A.W. and Brown, T.L. J. Am. Chem. Soc., 1971, 93, 1790-1791 (c) Falk, J.E. Porphyrins and Metalloporphyrins, Elsevier. Amsterdam, The Netherlands, 1964 (d) Lemberg, R. and Legge, J.W. Hematin compounds and _ Bile pigments, Iriterscience, New York, NY, 1949 [7] (a) Dolphin,D. (Ed.); The Porphyrins, Academic: New York Vol. III-V (b) Spiro,T.G. (Ed.); Biological Applications of Raman spectroscopy. Vol. 3, 1988 (c) Carey, P.R. Biochemical Applications of Raman and Resonance Ramon Spectroscopies, Academic: New York, 1982 (d) Hester, R.E. Chapter 4 in Raman spectroscopy, Szymanski, H.A. (Ed), Plenum, New York, NY, 1967 [8] (a) Asher, S. and Sauer, K. J. Chem. Phys., 1976, 64, 4115-4125 (b) Shelnutt, J.A.; O'Shea, D.C.; Yu, N.-T.; Cheung, L.D. and Felton, R. J. Chem. Phys., 1976, 64, 1156-1165 (c) Smith, D.W. and Williams, RJ.P. Structure and Bonding, Springer-Verlag, Berlin, Vol.7, 1970, 145 42 [9] (a) Williams, RJP. J. Thoeret. Biol., 1961, 1, 4 (b) Williams, RJ.P., in Hemes and Hemoproteins, eds. Chance, 8., Estabrook, R.W. and Yonetani, T., Academic Press, New York, 1966 [10] (a) Gouterman, M. J. Mol. Spec. 1961, 6, 138-163 (b) Gouterman, M.; Wagniere, OH. and Snyder, L.C. J. Mol. Spec. 1963, 11, 108-127 (c) Spellane, 5g; gigpterman, M.; antipas, A.; Kim, S. and Liu, Y.C. Inorg. Chem, 1980, 19, [11] (a) Oertling, W.A.; Salehi, A.; Chung, Y.C.; Leroi, G.E. Babcock, G.T. and Chang, C.K. J. Chem. Phys. 1987, 91, 5887-5898 (b) Sandusky, P.O., Oertling, W.A., Chang, C.K. and Babcock, G.T. J. Phys. Chem, 1991, 95, 4300-4307 [12] Wang, M.-Y.R. and Hoffman, B.M. J. Am. Chem. Soc., 1984, 106, 4235-4240 [13] Nakarnoto, K. Infrared and Ramon Spectra of Inorganic and Coordination Compounds 4th Ed., Wiley-Interscience Publication, 1985 [14] (a) Abe, M.; Kitagawa, T. and Kyogoku, Y. J. Chem. Phys. 1979, 69, 4526- 4534 (b) Gladkov, LL. and Solovyov, K.N. Spectrochemica Acta. 1986, 42A, 1- 10 [15] Dolphin, D., Muljiani, Z., Rousseau, K., Borg, D.C., Fajer, J. and Felton, R.H. Annals New York Academy of Sciences, 1973, 177-209 [16] (a) Boldt, NJ .; Donohoe, RJ.; Birge, RR. and Bocian, D.F. J. Am. Chem. Soc. 1987, 109, 2284-2298 (b) Czemuszewicz, R.S.; Macor, K.A.; Li, X.-Y.; Kincaid, LR. and Spiro, T.G. J. Am. Chem. Soc., 1989, 111, 3860-3869 [17] Lever, A.B.P. and Gray,, H.B., Eds., Iron Porphyrins, Parts I-III ,Addison Wesley Reading, Mass. 1983 [18] (a) Babcock, G.T. and Wikstrom, M. Nature, 1992, 356, 301-309 (b) Wikstrom, M. and Morgan, LE. J. Biol. Chem, 1992, 267, 10266-10273 [19] Watanabe, Y. and Groves, J.T. in The Enzymes, Chap. 9, Academic Press, Inc., New York, 1992 [20] (a) Balch, A.L; Chan, Y.-W. and Olmstead, M.M. J. Am. Chem. Soc. 1985, 107, 6510-6514 (b) Balch, A.L; Chan, Y.-W.; Olmstead, M.M. and Renner, M.W. J. Am. Chem. Soc. 1985, 107, 2393-2398 (c) Mizutani, Y., Watanabe, Y. and Kitagawa, T. J. Am. Chem. Soc. 1994, 16, 3439-3441 [21] Salehi, A; Oertling, W.A.; Babcock, G.T. and Chang, GK. J. Am. Chem. Soc. 1986, 108, 5630-5631 [22] (a) Kadish, K.M., Lin, X.Q. and Han, B.C. Inorg. Chem, 1987, 26, 4161- 4167 (b) Bottomley, LA. and Kadish, K.M. Inorg. Chem, 1981, 20, 1348—1357 43 [23] (a) Mu, X.H. and Kadish, K.M. Inorg. Chem, 1989, 28, 3743-3747 (b) Kadish, K.M. Prog. Inorg. Chem. 1986, 34, 435-605 (c) Chan, R.J.H., Su, Y.C. and Kuwana, T. Inorg. Chem, 1985, 24, 3777-3784 [24] (a) Yu, N.-T.; Kerr, B.A.; Ward, B and Chang, C.K. Biochemistry, 193, 22, 4534-4540 (b) Momenteau, M.; Mispelter, J .; Loock, B. and Lhoste, J .-M. J. Chem. Soc., Perkin Trans, 1985, 221-231 (c) Wijesekera, T.P.; Paine, J.B.III, and Dolphin, D. J. Org. Chem, 1988, 53, 1345-1352 [25] (a) Hashimoto, T. and Basolo, F. Comments Inorg. Chem, 1981, 1, 199- 206 (b) Hashimoto, T.; Dyer, R.L.; Crossley, J .E.; Baldwin, J .E. and Basolo, F. J. Am Chem. Soc., 1982, 104, 2101-2109 [26] Colman, J .P.; Gagne, R.R.; Reed, C.A.; Robinson, W.T. and Rodley, C.A. Proc. Natl. Acad. Sci., 1974, 71, 1326-1329 [27] (a) Collman, J .P.; Elliot, C.M.; Halbert, TR. and Tovrog, B.S. Proc. Natl. Acad. Sci.,U.S.A., 1977, 74, 18-22 (b) Ward, B.; Wang, C. and Chang, C.K. J. Am. Chem. Soc., 1981, 103, 5236-5238 [28] Antonini, E. and Brunori, M. Hemoglobin and Myoglobin in Their Reactions With Ligands, Elsevier, New York, 1971 [29] (a) Browett, W.R. and Stillman, MJ. Inorg. Chim. Acta., 1981, 49, 69-77 (b) Broweu, W.R. and Stillman, M.J. J. BioChim Biophys. Acta., 1981, 660, 1-7 [30] (a) Goff, HM. and Phillippi, M.A. J. Am. Chem. Soc., 1983, 105, 7567-7571 (b) Godziela, G.M. and Goff, H.M. J. Am. Chem. Soc., 1986, 108, 2237-2243 (c) Morishima, I.; T akamuki, Y. and Shiro, Y J. Am. Chem. Soc., 1984, 106, 7666- 7672 [31] Shimomula, E.T.; Phillippi, M.A.; Goff, H.M. Scholz, W.F. and Reed, CA. J. Am. Chem. Soc., 1981, 103, 6778-6780 [32] (a) Barkigia, K.M.; Spaulding, L.D. and Fajer, J. Inorg. chem, 1983, 22, 349- 351 (b) spaulding, L.D.; Eller, P.G.; Bertrand, LA. and Felton, R.H. J. Am. Chem. Soc., 1974, 96, 982-987 [33] (a) Pauling, L. The Nature of the Chemical Bond, Comel University Press, Ithaca, 1960 (b) Pimentel, G.C. and McClellen, AL. The Hydrogen Bond, W.H. Freeman and Co., New York, 1960 [34] (a) Branden, C. and Tooze, J. Introduction to Protein Structure, Garland Publishing, New York, 1991 (b) Atwood, J .1... Hamada, F., Robinson, K.D., Orr, G.W. and Vincent, R.I. Nature, 1991, 349, 683-684 (c) Jeffrey, G.A. and Saenger, W. Hydrogen Bonding in Biological Structures, Springer-Verlag, Berlin Heidelberg New York, 1991 44 [35] For a current set of review articles on hemoglobin, myoglobin and corresponding model compounds see for example Chemical Reviews, May 1994, Volume 94, Number 3 [36] (a) Collman, J .P.; Brauman, J .I.; Iveson, B.L.; Sessler, J .L.; Morris, R.M. and Gibson, Q.H. J. Am. Chem. Soc., 1983, 105, 3052-3064 (b) Lavalette, D.; Tetreau, C.; Mispelter, J .; Momenteau, M. and Lhoste, J .-M. Eur. J. Biochem, 1984, 106, 5132-5143 (c) Traylor, T.G.; Koga, N .; Deardurff, L.A.; Mitchell, M. and Stynes, J .A. J. Am. Chem. Soc., 1984, 106, 5132-5143 (d) Traylor, T.G.; Tsuchiya, S.; Campbell, D.; Mitchell, M.; Stynes, D. and Koga, N. J. Am. Chem. Soc., 1985, 107, 604—614 (e) Springer, B.A.; Bgeberg, K.D.; Sligar, S.G.; Rohlfs, R.J.; Mathews, AJ. and Olson, J .S. J. Biol. Chem, 1989, 264, 3057-3060 [37] (a) Varotsis, C.; Woodruff, W.H. and Babcock, G.T. J. Am. Chem. Soc. 1989, 111, 6439-6440 (b) Han, S.; Ching, Y.-c. and Rousseau, D.L. Proc. Natl. Acad. Sci.,U.SA. 1990, 87, 2491-2495 (c) Varotsis, C. and Babcock, G.T. Biochem 1990, 29, 7357-7362 (d) Varotsis, C.; Woodruff, W.H. and Babcock, G.T. J. Biol. Chem 1990, 265, 11131- [38] (a) Varotsis, C.; Zhang, Y.; Appelman, RH. and Babcock, G.T. Proc. Natl. Acad. Sci.,U.SA. 1992, 90, 237-241(b) Ogura, T.; Takahashi, S.; Hirota, S; Shinzawa-Itoh, K; Yoshikawa, S.; Appelman, RH. and Kitagawa, J. J. Am. Chem. Soc. 1993, 115, 8527-8536 [39] (a) Garrett, A.W. and Zwier, T.S. J. Chem. Phys. 1992, 96, 9710-9718 (b) Gotch, AJ. and Zwier, T.S. J. Chem. Phys. 1990, 93(10), 6977-6986 [40] Tuchen, E. and Woodward, C. Biochemistry, 1987, 26, 1918-1925 [41] Kitagawa, T.; Ondrias, M.R.; Rousseau, D.L.; lkeda-Saito, M. and Yonetani, T. Nature 1982, 26, 869-871 [42] (a) Smith, M.L.; Ohlsson, P.-I. and Paul, KG. FEBS Letters 1983, 163, 303- 305 [43] Satterlee, JD. and Erman, J .E. J. Am. Chem. Soc. 1984, 106, 1139-1140 [44] For example see Chapter 8 in Peiffer, W.E., A. Ph.D. Dissertation, Michigan State University, 1992 [45] (a) Oertling, W.A.; Kean, R.T.; Wever,R. and Babcock, G.T. Inorg. Chem, 1989, 29, 2633-2645 (b) Groves, J .T .; Huashalter, R.C.; Nakamura, M.; Nemo, TE. and Evans, B.J. J. Am. Chem. soc, 1981, 103, 2884-2886 (c) Groves, J .T.; Nemo, TE. and Myers, R.S. J. Am. Chem. soc., 1979, 101, 1032-1033 ((1) Chang, GK. and Kuo, M.-S. J. Am. Chem. soc., 1979, 101, 7613-7615 45 [46] (a) Pettersen, R.C. and Alexander, L.E. J. Am. Chem. soc., 1968, 90, 3873- 3875 (b) Buchler, J .W.; Puppe, L.; Rohbock, K. and Schneehage, H.H. Ann. NY. Acad. Sci., 1973, 206 116 (c) Su, O.Y.; Czemuscewicz, R.S.; Miller, LA. and Spiro, T.G. J. Am. Chem. Soc. 1988, 110, 4150-4157 [47] Procyk, A.D., Kim, Y., Schmidt, 13., Fonda, H.N., Chang, C.K., Babcock, G.T. and Bocian, D.F. J. Am. Chem. Soc. 1992, 114, 6539-6549 [48] Zhang, H.; Schmidt, E.; Wu, W.; Chang, G.T. and Babcock, G.T. Chem. Phys. Lett., Submitted, 1994 [49] Fetter, J ., Schmidt, E., Babcock, G.T. Fergusson-Miller, S. Proc. Nat. Acad. Sci., Submitted 1994 CHAPTER 2 Preparation and Characterization of 5 and 6 Coordinated (C0)ComOEP(ClO4') Complexes, Electron Absorption and Vibrational Spectroscopies and Normal Coordinate Analysis. 2.1 Overview When C0(g) was added to solutions of the cobaltous porphyrin 1c cation radical [Co(II)OEP-]C104, prepared from oxidation of CoOEP by AgClO4 in anhydrous CH2C12, room temperature binding at the metal center occurred. Two distinct products were formed, [(C0)Co(III)OEP]CIO4 and [(C0)2Co(III)OEP]ClO4. These compounds exhibit Soret maxima at 366 and 414 nm, and form with P1/2 values of 36 :l: 3 and 4000 :t 300 torr of CO, respectively. Isosbestic points in the optical absorption spectra occurred at 368 nm for binding of the first CO and at 385 nm during binding of the second ligand. FTIR and resonance Raman (RR) spectra of [(C0)Co(III)OEP]ClO4 reveal vibrations at 2110 and 441 cm"1 that shift to 2060 and 435 cm'1 upon substitution of 13C() for the natural abundance CO. Isotope sensitive vibrations of [(C0)2Co(III)OEP]ClO4 were measured at 2137 and 468 cm’l. These relatively high C-O stretching frequencies of 2137 and 2110 cm'1 are suggestive of relatively weak metal dtt-ligand 1t* backbonding, resulting from the oxidation of Co(II) to Co(III). RR spectra before the addition of CO were used to characterize 46 47 the cobaltous porphyrin 1t cation radical, and after the addition of CO, RR spectra obtained by using 363.8 nm excitation were used to confirm the occurrence of the five-coordinate Co(IIl) porphyrin complex. This complex, along with halide ligated analogs, displays some structure sensitive frequencies that suggest that an unusual distortion of the porphyrin core occurs in CH2C12 solution. RR spectra obtained by using 413.1 nm excitation after the addition of CO were used to identify the second product as a typical six-coordinate cobaltic porphyrin. 2.2 Introduction Synthetic cobalt porphyrin complexes have been used as models for oxygen transport proteins [1]. Low temperature binding of diatomics such as 02, CO, and NO to conventional cobaltous porphyrins has been achieved [2], with ligation perhaps best established by vibrational spectroscopy [3]. Binding of dioxygen has received the most attention and matrix isolation techniques have been used to facilitate these studies [3]. As is typical of diatomic ligands, the 0-0 stretching frequency is modified by bonding to the metal and can usually be detected by IR [4,5], whereas resonance enhancement of the lower frequency Co-02 vibrations make these frequencies readily available from Raman measurements [6,7]. Thus, it is useful to use both techniques in order to establish all metal-ligand vibrational frequences. Vibrational assignments are necessarily verified by isotopic substitutions on the ligand atoms, which affects the vibrational frequencies in these simple systems in a relatively straightforward manner [8]. Use of specialized protected porphyrins allow oxy cobaltous adducts in solution at room tempterature to be studied by RR [9]. Possible protein interactions can be modeled and their effects on vibrational frequencies assessed in this manner [10]. Some interesting studies of vibrational coupling between bound dioxygen vibrations and vibrational modes of trans ligands and solvent 48 have been done with oxy cobalt porphyrins [11]. Cobalt porphyrins reconstituted into myoglobin and hemoglobin have been characterized by resonance Raman (RR) measurements [12]. Oxy derivatives of reconstituted cobalt myoglobin are stable at room temperature and RR measurements of these complexes originally revealed the 0-0 stretching frequency [6]; indeed, the 0-0 stretching vibration is not easily observed in oxy Fe myoglobin by Raman or IR [13] measurements. Thus, many RR studies of oxygen transport proteins have used these oxy cobaltous adducts [14]. Oxidized derivatives of cobalt porphyrins have been looked upon as models for intermediates in the enzymatic functions of catalases, peroxidases and cytochromes P450 [15]. During the catalysis cycle, the heme cofactors of these enzymes are oxidized by two electrons, usually at both the metal center and the porphyrin macrocyle to yielding an oxoferryl porphyrin 1r. cation radical, [O=Fe(IV)P-]+. Like the heme cofactor, cobalt porphyrins readily lose one electron from each of the metal and the ring macrocycle. In particular, when cobaltous octaethylporphyrin CoOEP is oxidized by two electrons to [Co(III)OEP-]2+, two distinct species are possible, one grey and one green in color [16,17]. These two distinct uv-visible absorbance signatures were originally thought to be indicative of the electronic states, 2A1“ or 2A2“, but later Raman [18], EPR and ENDOR [19,20] work by our group as well as NMR [21] and Raman [22] work by others, clearly showed that these two forms both have a 2A1“ ground electronic state. Rather, most likely differences in macrocycle conformation cause variations in the porphyrin nitrogen donation to the metal, which gives rise to the distinct spectral types [20]. Loss of one electron from CoOEP can result in a number of possible products, depending primarily on the presence of coordinating ligands. Salehi et al. [23] showed that one-electron oxidation of this compound by AgClO4 in 49 scrupulously dry CH2Cl2 results in ring-centered oxidation yielding a pale, bluish-gray [Co(II)OEP-1004. This compex is sensitive to the presence of even weakly ligating species (eg. H20). Any ligand interactions with the C001) (a d7 complex) will cause an intramolecular electron transfer and "push" the odd electron in the metal d22 orbital into the electron deficient ring system. We have observed that reactions of the type [Co(II)OEP-1004 + L --> [LCo(III)OEP]ClO4 [Co(II)OEP-1004 4» 2L --> [L2Co(IlI)OEP]ClO4 occur very easily. These adducts are at the same formal oxidation level, yet each has a distinct spectral signature. The cobaltous 1t cation radical has an absorbance maximum at 377 nm, quite different from either the grey or green cobaltic 1t cation radicals [16,24]. The five-coordinate cobaltic complex is easily made when L is a halide anion and exhibits two somewhat diffuse optical absorbances at ~373 and 546 nm, whereas the spectra of six coordinate complexes display sharper features at ~410, 525 and 558 nm for fairly weak ligands, with these features shifting to the red as the ligand strength increases [23,24]. It is not uncommon for samples of the one electron oxidized adduct to display heterogeneity and many spectra in the literature are due to a mixture of these species. In contrast to the extensive investigation of dioxygen complexes of cobalt porphyrins described above, there are relatively few reports of carbon monoxy cobaltous complexes [2,3] and none at room temperature. Interestingly, Kadish and coworkers have recently presented evidence suggesting that CO will bind to cobalt porphyrins at room temperature when a one-electron oxidation of a cobaltous TPP (tetraphenylporphyrin) or OEP is carried out electrochemically under a CO atmosphere [25,26]. The result is a COCo(III) adduct. Preceeding this work, there was only a single report of a cobaltic CO complex [27], and it is 50 considered somewhat odd for CO to bind metal ions in’ higher oxidation states due to the 1: acid nature of the ligand [28]. If oxidation of CoOEP under CO initally produces a cobaltous 1t cation radical with subsequent CO binding, then CO may simply serve as the ligand L in this case, analogous to the reactions depicted above. In the present work, we have prepared [Co(II)OEP-lClO4 and then introduced CO in a separate step. Here we report conclusive evidence of CO binding to Co(III), namely both Co-CO and CoCEO stretching frequencies, verified with isotopic substitution. Furthermore, we demonstrate that both five coordinate [(CO)Co(III)OEP]C104 and six coordinate [(C0)2Co(III)OEP]ClO4 can be formed, depending on C0 pressure, and we measure formation equilibrium constants for these species. Our analysis reveals that the earlier work resulted in formation of predominantly the dicarbonmonoxy derivative, rather than the 'monocarbonmonoxy adduct, as was reported [26]. 2.3 Experimental Porphyrins were synthesized [29,30] and metals inserted according to established methods [31]. Oxidations were carried out by adding an excess of anhydrous AgClO4(s) to the cobalt porphyrin solution and stirring for 1 to 3 hours in an evacuated cell. CO(g) was typically bubbled into the sample through a syringe, while a second needle was used as a vent to expel excess gas. For the CO titration, the sample was prepared in the usual way in dry CHzClz. CO was introduced with a calibrated gas tight syringe. Before each injection, the pressure within the gas tight syringe was allowed to equilibrate to atmospheric pressure, estimated at 760 Torr. The titration vessel was 135 mL capacity and the partial pressure of CO was calculated by starting under vacuum, then systematically injecting into the vessel a known volume of CO gas. Ideal 51 gas behavior was assumed and the partial pressure was calculated from (Vinjected/Vvessel)"‘760 Torr. FI'IR requires highly concentrated samples. Due to the limited solubility of CoOEP in CH2C12 these concentrations were unattainable. However, dimethyl- 2,4,6,8-tetramethyl-3,7-dioctylporphine-1,5-dipropanoate cobalt(II)porphyrin [30] (CngP) was sufficiently soluble in CH2C12 so that FI'IR samples could be made. To ensure that CngP had redox and ligation properties similar to those of CoOEP, we used optical absorption spectroscopy to follow the oxidation and CO titration of CngP. FTIR samples were prepared by placing 1 mg of cobalt porphyrin and anhydrous AgClO4 into a vial sealed with a rubber septum and flushed with N2(g). Freshly distilled methylene chloride (1 mL) was injected into the vial with a Hamilton gas tight syringe. Deuterated solvent was used as supplied by Cambridge Isotope Laboratories. After oxidation was complete, the sample was transferred by a gas tight syringe into a 50 um NaCl cavity cell. The cavity cell was in a holder sealed with a rubber septum. Carbon monoxide was bubbled into the sample via a syringe needle inserted to the groove of the cavity cell. A second needle was used as a vent to expel excess CO. Solution infrared spectra were recorded on a Nicolet IR/42 FTIR spectrometer. Raman spectra were obtained by using a 90° scattering geometry and were measured with either a Spex 1401 equipped with a PMT or a Spex 1877 monochromator equipped with an OMA III diode array detector. Laser emissions at 363.8 and 413.1 nm were from a Coherent Innova 200 argon ion laser and a Coherent Innova 90 krypton ion laser, respectively. Optical absorption spectra, obtained by using a Perkin-Elmer Lamda 5 spectrometer, were recorded prior to and following Raman studies to confirm sample integrity. Anaerobic quartz cuvettes were used for both the Raman and optical absoption measurements. 52 Normal coordinate calculations were carried out on an XS-ma 24, 4000 SGI machine using standard GF-matrix methods [32,33]. Initial force constants were chosen from Jones et al. [34]. Force constants were optimized by the method of McIntosh and Michaelan [33]. 2.4 Results 2.4.1 Optical Spectroscopy Uv-vis absorbance spectra depicting the sequential oxidation and ligand binding of CoOEP are shown in Figure 1. The starting material (solid line) is oxidized by one-electron in anhydrous CH2C12 to give the cobaltous 1t cation radical [Co(II)OEP-lClO4 (dashed line). The spectrum of this species displays a Soret maxima at 377, as reported in earlier measurements [23 ,24]. When gaseous CO is introduced above the solution to a pressure of approximately 1 atrn, a split Soret band with maxima at 366 and 414 nm develops, as in Figure 1c (dotted line). Also, typical Q-band absorptions at approximately 528 and 557 nm replace the broad of the it cation radical, suggesting that the porphyrin ring has been reduced by one electron to form the ring neutral species. The relative intensities of the two absorptions in the split Soret band of Figure 1c are dependent on C0 pressure and titration with CO yields spectra with well defined isosbestic points (see below). This suggests that the bands at 366 and 414 nm arise from distinct complexes. Thus, initial oxidation results in a cobaltous 1t cation radical and subsequent introduction of CO seems to yield two distinct, possibly nonradical complexes. 2.4.2 Resonance Raman spectroscopy By using laser lines in resonance with each of the two absorbance maxima, Raman measurements can provide selective enhancement of vibrational spectra 53 Figure 2.1 a) (—) Neutral ConOEP in CH2C12 b) (---) One electron oxidized CoIIOEP+-(c104-) in CH2C12 under an N2 atmosphere. c) (~---) One electron oxidized ConOEP+- (ClO4') in CH2C12 under a CO atmosphere. 54 A AA! “ Q‘Nt 600 l 500 WAVELENGTH (nm) 400 300 1.0-"- 091- 0.8”- o 7 -l- 0.4 ~- - q 5. 0 p 1 6. o moz._..m2m._.z_ m>uh<4m¢ T T f Y 1000 1200 1400 1600 060 RAMAN SHIFT (cm-1) j # ‘M 600 _~r_ 200 58 Table 1. Structural sensitive vibrational frequencies (cm' 1) observed in resonance Raman spectra obtained at the given excitation wavelengths (nm). ConOEP [CoHOEP-1 [(C0)2C0(111)0EP1C104 (2) 366 nm 414 nm [Co(II)OEP-lClO4 + 2C0 <=> [(C0)2Co(III)OEP]ClO4 (3) We show below that we can neglect (3) for this system. Vibrational analysis of the cobalt-carbonmonoxy vibrations provides further evidence for this scheme and is also presented below, followed by determination of the equilibrium constants for ( 1) - (3). 2.5 Discussion 2.5.1 Analysis of Cobalt-CO vibrations. 2.5.1.1 Group theoretical analysis The five-coordinate complex [(CO)Co(III)OEP]ClO4 belongs to the C4v point group and a linear Co-CI-EO unit is expected to give rise to four vibrations given by: 2A1 + E. The two vibrations of A1 symmetry are totally symmetric stretching motions while the E vibrations are a degenerate pair of linear bends. All vibrations are expected to be both Raman and IR active [38]. The linear bends, being degenerate, should occur at the same frequency and give rise to a single feature in the 200 - 600 cm'1 region. One stretch is expected to be primarily v(CEO) occuring around 2100 cm"1 and the other v(Co-C) at around 300-600 cm'l. There is little mixing of these stretching coordinates because of the large difference in frequencies. 7O begins to resolve. The two clearly defined isosbestic points provide credibility for our suggestion of two ligation events. Thus, the pressure dependence of the absorbance spectra suggests the formation of two complexes that are in equilibrium as shown by the following: [Co(II)OEP-]CIO4 + CO <=> [(CO)Co(III)OEP]ClO4 (1) 377 nm 366 nm [(CO)Co(III)OEP]ClO4 + CO 4:) [(C0)2Co(III)OEP]ClO4 (2) 366 nm 414 nm [Co(II)OEP-]ClO4 + 2C0 <=> [(CO)2Co(IH)OEP]ClO4 (3) We show below that we can neglect (3) for this system. Vibrational analysis of the cobalt-carbonmonoxy vibrations provides further evidence for this scheme and is also presented below, followed by determination of the equilibrium constants for (1) - (3). 2.5 Discussion 2.5.1 Analysis of Cobalt-CO vibrations. 2.5.1.1 Group theoretical analysis The five-coordinate complex [(CO)Co(III)OEP]ClO4 belongs to the C4v point group and a linear Co-CEO unit is expected to give rise to four vibrations given by: 2A1 + E. The two vibrations of A1 symmetry are totally symmetric stretching motions while the E vibrations are a degenerate pair of linear bends. All vibrations are expected to be both Raman and IR active [38]. The linear bends, being degenerate, should occur at the same frequency and give rise to a single feature in the 200 - 600 cm'1 region. One stretch is expected to be primarily v(CEO) occuring around 2100 cm'1 and the other v(Co-C) at around 300-600 cm'l. There is little mixing of these stretching coordinates because of the large difference in frequencies. 71 The six-coordinate complex [(C0)2Co(III)OEP]ClO4 most likely belongs to the D41; group. A linear OEC-Co-CEO unit is expected to generate ten vibrations given by: 2A1g + 2A2“ + 2Eg + Eu. The A type modes are stretches and the E type modes are degenerate linear bends. Gerade vibrations are Raman active and ungerade vibrations are IR active [38]. The frequency regions in which the vibrations are expected are similar to the C4v case above. For example, a totally symmetric (A1g) Raman active mode <--OECCoCEO--> is ' expected around 2100 cm'l, while another, corresponding to symmetric stretching of the metal-carbon bonds <--OC-Co-CO--> , is expected in the low frequencies. The situation is similar for IR active A21. antisymmetric stretching modes. 2.5.1.2 Vibrational spectroscopy Table II collects the observed metal-ligand vibrational frequencies, calculated frequencies obtained by using FG matrix methods, and our vibrational normal mode assignments for both the five and six-coordinate complexes. The appearance of the 2110 cm'1 vibration in both the RR and IR data is strong support for a five-coordinate C4v complex. The assignment of the other high frequency IR feature to the asymmetric coupled CEO stretch of the six- coordinate complex is reasonable. Our analysis predicts the presence of another Raman active high frequency mode ~2300 cm'1 corresponding to the symmetric coupled CEO stretch for this six-coordinate species, but we did not observe it by using 413.1 nm excitation. This is not surprising because the resonance enhancement of these modes is typically weak, and, because Raman intensity is proportional to the fourth power of the absolute scattered frequency, detection of this mode by using 413.1 nm excitation is expected to be more difficult than by using 363.8 nm excitation (Figure 4). 72 Based on the number of d electrons in the metal (6) and the number of 7t* electrons in the ligand (0), the geometry of the Co(III)-C-O is expected to be linear [8]. Yu et al. [39] report no RR enhancement of metal-ligand bending vibrations for complexes known to have linear M-C-O geometries. Similarly, we observed no bending motions in the RR spectra. In the course of doing our normal coordinate calculations, we were able to fit our observed frequencies only when linear Co-C-O geometries were input. Thus, CO most likely binds to Coin a linear fashion in these complexes. The C—O stretching frequencies observed, 2137 cm'1 for the six-coordinate and 2110 cm"1 for the five coordinate complex, reflect relatively weak metal du- ligand 1t* backbonding. This reflects the decreased electron density at the Co(III) due to the high oxidation state. Metals in lower oxidation states are able to more strongly donate electrons back into the 1t* orbital of the ligand, causing a greater decrease in the C-0 stretching frequency [40]. For example, C-O frequencies (cm‘l) observed for some M(II) porphyrin complexes include: (CO)2CoTPP, 2077; (CO)2FeTPP, 2042; (CO)2RuTPP, 2005 [3]. The 2137 cm‘1 that we observe for [CO)2Co(III)OEP]ClO4 is hardly lowered from the frequency of free CO ~2155 cm'l, indicating that the bonding interaction is dominated by a donation from the ligand to the metal in this M(III) complex. Judging from our observed frequecies, the backbonding in the six-coordinate complex is even weaker than in the five- coordinate complex because in the six-coordinate complex the metal donation must be split between two ligands. Comparison of the C-0 frequency for (CO)2FeTPP, 2042 cm], to that of (CO)FeTPP, 1973 cm'l, illustrates the same trend [26]. 73 2.5.2 Calculation of equilibrium constants and Pl/z values for reactions (1) - (3). Assuming that the system follows Henry's law, relating the partial pressure of a gas above a solution to its solution concentration, we can write equilibrium constant expressions for (l) and (2): = [(C0)Co(III)0EP]+ K ' [Co(II)OEPTPw = [(C0)LCo(III)OEP]* ’ [(C0)Co(III)OEP]*Pw The true concentration equilibrium constants can be obtained by multiplication of K1 and K2 by the appropriate Henry's constant for CO in CH2C12. K3 (for the third reaction) is the product K1K2 and is negligible for this system. K1 and K2 can be determined from Figure 6 by using the following [41]: l l l 1 * A-A.=K.(A.-A.> aim-A.) where A represents the absorbance of the system at a particular partial pressure of CO, A0 is the initial absorbance of the reactant species, K1 is K1 or K2, A... is the final absorbance of pure product species, and FCC is the partial pressure of CO. Thus, a plot of l/(A-Ao) vs. 1/Pco will give a line. A... is determined from the intercept and K1 from the slope. Because only negligible formation of the six- coordinate complex occurs before reaction (1) has gone to completion, we can use the expression above to determine either K1 or K2 [42]. Figures 7 and 8 show such plots for reactions (1) and (2), respectively. P1 /2, represents the pressure of CO when the reaction has proceeded halfway, is given by the 74 reciprocal of the equilibrium constant, P1 /2 = (Ki)'1. From Figures 7 and 8 we obtain P1/2 estimates of 36 :l: 3 and 4000 :l: 300 torr for reactions (1) and (2), respectively. 2.5.2 Comparison of [(CO)Co(III)OEP]ClO4 to BrCo(III)OEP and other complexes. The optical absorbance spectrum of [(CO)Co(III)OEP]ClO4 is similar to that of BrCo(III)OEP [24]. Both display dramatically blue-shifted Soret absorbances (366 and 373 nm, respectively, in CHzClz) compared to that of typical metalloporphyrin complexes. ClCo(III)OEP and ICo(III)OEP also display this blue-shifted Soret maximum [43]; however, we are aware of few other metalloporphyrins with this peculiar spectral signature. The ~370 nm Soret band is odd in two regards. First, oxidation of Co(II) to Co(III) normally causes a red- shift in porphyrin compounds [44]. (This is of course different for iron porphyrin complexes, which are not germane to the discussion of this topic.) Thus, the Soret maximum of cobaltic OEP compounds are expected to be red-shifted relative to that of Co(II)OEP (391 nm), as are the Soret absorbances of the six- coordinate adducts. In the absence of spin or oxidation-state changes,“5 ligation of a single axial ligand normally causes a red-shift in the spectrum of the metalloporphyrin, with a second ligand simply causing an additional red-shift. For example, spectral changes of Cu(II) porphyrins [46] upon ligation display the typical red-shifted trend. Indeed, this is what Kadish and coworkers[27] observed for Co(III)TPP+ (TPP = tetraphenylporphyrin) complexes upon sequential ligation of CH3CN. Thus, the formation of this five-coordinate complex displaying the blue-shifted Soret band appears to be unique to cobaltic complexes of specific porphyrins in solution. Interestingly, (CN)Co(III)OEP is reported to exhibit distinct Soret maxima at 372 and 421 nm [47], while (CN)Co(III)protoporphyrin IX reconstituted in horseradish peroxidase displays a 75 Soret maximum at 436 nm [48]. The 372 and 421 nm maxima most likely represent two different conformations of the five-coordinate cobaltic cyano complex. The 372 nm band most likely arises from this rather unique configuration and the 421 nm band from a more conventional geometry. The 436 nm maximum represents the effect of a sixth ligand (histidine, and possibly other influences) from the protein. It is well known that RR frequencies are much more informative of structural properties of metalloporphyrins [35] than optical absorbance maxima. Spaulding et al. [49] first recognized the linear correlation between the frequency of V19, an anomalously polarized (ap) mode, and the porphyrin center to nitrogen distance. Similar correlations were demonstrated for virtually all high frequency vibrations [50-52]. Metalloporphyrins exhibiting ruffled or domed geometries, for example ferrous complexes, were shown to display V10 and V19 frequencies which were significantly lower than expected based on their core sizes [49-52]. Not surprisingly, BrCoOEP was noted to display Raman frequencies that uniquely deviated from these correlations. The BrCoOEP frequencies most sensitive to core-size are: 1657 (V10), 1599 (V2), 1575 (V11), 1572 (V19) and 1514 (V3). In particular, the difference in the V10 (dp) and V10 (dp) and V19 frequency of metalloporphyrins is typically 54 i 4 cm'l, whereas in BrCoOEP it is 82 cm’1 [49]. Comparison of the spectra of Co(II)OEP and BrCo(III)OEP and [(MeOH)Co(III)OEP]Br [24] reveals that all of the RR frequencies between 1300 - 1700 cm'1 are essentially identical for the three complexes, with the exception of V10 and V19. In the spectra of BrCo(III)OEP the V10 frequency is elevated and the V19 frequency lowered from typical values. We have speculated that this results from a severe structural perturbation in solution [24], and doming of the porphyrin with the metal ion significantly out of the plane has been suggested [49,53]. The elevation of V10 is puzzling, however, as both ruffling and doming 76 distortions are expected lower both of these frequencies by ~15 cm'1 [52]. We have not been able to obtain crystals of this compound suitable for x-ray crystallography. . Although the V3 frequency for [(CO)Co(III)OEP]ClO4 is somewhat less than that of BrCoOEP, the similarity of the other vibrational frequecies and the Soret maxima of these two complexes attests to shared structural features. We have noted [24] that a u-nitrido dimer (OEPFe)2N displays some similar spectral features [54] to the bromide complex and have speculated that the severely ruffled porphyrin core of this Fe complex [55] may model the Co(III) complexes studied here. Shelnutt et al. [56] have pointed out that aggregates of some metallouroporphyrins display similarly blue-shifted Soret absorbances. Our dilution studies show no concentration dependence of the Soret absorbance of BrCoOEP, as would be expected for a dimer or aggregate. At any rate, the solution structure of these five-coordinate Co(III) complexes most likely involves some distorted porphyrin conformation. This conformation perhaps makes the ligation of the sixth ligand more difficult, keeping the equilibrium constant K2 relatively low. 77 REFERENCES [l] (a) Stynes, D.V., Stynes, H. C., James, B. R. and Ibers, J. A. J. Am Chem. Soc. 1973, 95, 1796-1801. (b) Collman, J. P., Brauman, T. R. H. and Suslick, K. S. Proc. Natl. Acad. Sci. USA , 1976, 73, 3333-3337 (b) Jones, R. D., Summerville, D. A. and Basolo, F. Chemical Reviews 1979, 79, 139-179 (c) Liu, H. -Y., Abdalmuhdi, 1., Chang, C. K. and Anson, F. C. J. Phys. Chem. 1985, 89, 665-670 [2] (a) Wayland, B. B. and Mohajer, D. J. Am. Chem. Soc. 1971, 93, 5295-5296 (b) Wayland, B. B., Minkiewicz, J. V. and Abd-Ehnageed, M. E. J. Am. Chem. Soc. 1974, 96, 2795-2801 (c) Wayland, B.B.; Mehne, L. F. and Swartz, J. J. Am. Chem. Soc. 1978, 100, 2379-2149 [3] Kozuka, M. and Nakamoto, K. J. Am. Chem. Soc. 1981, 103, 2162-2168 [4] Urban, M. W., Nakamoto, K. and Kincaid, J. lnorganica Chim. Acta 1982, 61, 77-81 [5] Nakamoto, K. Paeng, I. R., Kuroii, T., Isobe, T. and Oshio, H. J. Mol. Struc. 1988, 189, 293-300 [6] Tsubaki, M. and Yu, N.-T. Proc. Natl. Acad. Sci. USA 1981, 78, 3581-3585 [7] Bajdor, K., Nakamoto, K and Kincaid, J. J. Am. Chem. Soc. 1983, 105, 678- 679 [8] for a review see: Yu, N. -T., & Kerr, E. A. in Biological Applications of Raman Spectroscopy, vol 3, 1988 Spiro, T. G., Ed., p39-95 [9] Mackin, H. C., Tsubaki, M. and Yu, N. -r. Biophys. J. 1983,41, 349-357 [10] (a) Odo, J ., Irnai, H., Kyuno, E. and Nakamoto, K. J. Am. Chem. Soc. 1988, 110, 742-748 (b) Proniewicz, L. M., Odo, J ., Gobral, J ., Chang, C. K. and Nakamoto, K. J. Am. Chem. Soc. 1989, 111, 2105-2110 (c) Proniewicz, L. M., Bruha, A., Nakamoto, K., Kyuno, E. and Kincaid, J. R. J. Am. Chem. Soc. 1989, 1 1 1, 7050-7056 [1 l] (a) Bajdor, K., Kincaid, J. R. and Nakamoto, K. J. Am. Chem. Soc. 1984, 106, 7741-7747 (b) Kincaid, J. R., Proniewicz, L. M., Bajdor, K., Bruha, A. and 78 Nakamoto, K. J. Am. Chem. Soc. 1985, 107, 6775-6781 (c) Proniewicz, L. M. and Kincaid, J. J. Am. Chem. Soc. 1990, 112, 675-681 [12] (a) Woodruff, W. H., Spiro, T. G. and Yonetani, T. Proc. Natl. Acad. Sci. USA 1974, 71, 1065-1069 (b) Woodruff, W. 11., Adams, D. H., Spiro, T. G. and Yonetani, T. J. Am. Chem. Soc. 1975, 97, 1695-1698 [13] (a) Barlow, C. H.; Maxwell, J. C., Wallace, W. J. and Caughey,W. S. Biochem. Biophys. Res. Commun. 1973, 55, 91- 95 (b) Potter, W. T., Tucker, M. P. ., Houtchens, R. A. and Caughey, W. S. Biochemistry 1987, 26, 4699-4707 [14] (a) Kitagawa, T., Ondrias, M. R., Rousseau, D. L., Ikeda-Saito, M. and Yonetani, T. Nature 1982, 298, 869-871 (b) Thompson, H. M., Yu, N.-T. and Gersonde, K. Biophys J. 1987, 87, 289-295 (c) Bruha, A. and Kincaid, J. R. J. Am. Chem. Soc. 1988, 110, 6006-6014 [15] (a) Frew, J. E. and Jones, P. In Advances in Inorganic and Bioinorganic Mechanism; Academic Press: New York, 1984; Vol.3, pp175-212 (b) Groves, J. T. In Cytochrome P-450: Structure, Mechanism and Biochemistry; Ortiz de Montellano, P., Ed.; Plenum Press: New York, 1986; Chapter 1 [16] Dolphin, D., Fonnan, A., Borg, D. C., Fajer, J. and Felton, R. H. Proc. Natl. Acad. Sci. USA 1971, 68, 614-618 [17] Fuhrhop, J.-H., Kadish, K. M. and Davis, D. G. J. Am. Chem. Soc. 1973, 95, 5 140-5147 [18] Oertling, W. A., Salehi, A., Chang, C. K. and Babcock, G. T. J. Phys. Chem. 1989, 93, 1311-1319 [19] Sandusky, P. O., Salehi, A., Chang, C. K. and Babcock, G. T. J. Am. Chem. Soc. 1989, 11 1, 6431-6439 [20] Sandusky, P. O., Oertling,W. A. ,Chang, C. K. and Babcock, G. T. J. Phys. Chem. 1991, 95, 4300—4307 [21] (a) Morishima, 1., Takamuki, Y. and Shiro, Y. J. Am. Chem. Soc. 1984, 106, 7666-7672 (b) Godziela, G. M. and Goff, H. M. J. Am. Chem. Soc. 1986, 108, 2237-2243 [22] Czemuszewicz, R. S., Macor, K. A., Li, X. Y., Kincaid, J. R. and Spiro, T. G. J. Am. Chem. Soc. 1989, 111, 3860-3869 [23] Salehi, A., Oertling, W. A., Babcock, G. T. and Chang, C. K. J. Am. Chem. Soc. 1986, 108, 5630-5631 [24] Oertling, W. A., Salehi, A., Chung, Y. C., Leroi, G. E., Chang, C. K. and Babcock, G. T. J. Phys. Chem. 1987, 91, 5887-5898 79 [25] Mu, X. H. and Kadish, K. M. Inorg. Chem. 1989, 28, -3743-3747 [26] Hu, Y., Han, B. C., Bao, L. Y., Mu, X. H. and Kadish, K. M. Inorg. Chem. 1991, 30, 2444—2446 [27] Herlinger, A. W. and Brown, T. L. J. Am. Chem. Soc. 1971, 93, 1790-1791 [28] Jones, L. H., McDowell, R. S. and Goldblatt, M. Inorg. Chem. 1969,11, 2349-2363 [29] Wang, C. B. and Chang, C. K. Synthesis 1979, 548-459 [30] Eaton, S. 8.; Eaton, G. R. and Chang, C. K. J. AmChem. Soc. 1985,107, 3177-3184 [31] Falk, J. E in Porphyrins and Metalloporphyrins; Elsevier: New York, 1964, p798 [32] Wilson, W. E. B., Jr.; Decius, J. C. and Cross, P. C. in Molecular Vibrations, McGraw-Hill: New York, 1955; Chapter 4 [33] McIntosh, D. F. and Michaelan, K. H. Canadian J. Spectr., 1979,24, 1-10; 35-40; 65-74 [34] Jones, L. H.; McDowell, R. S. and Goldblatt, M. Inorg. Chem, 1968,8, 2350-2363 [35] Spiro, T. G., Czemuszewicz, R. S. and Li, X. -Y. Coord. Chem. Rev. 1990, 100, 541-571 [36] Abe, M., Kitagawa, T., and Kyogoku,Y. J. Chem. Phys. 1978, 69,4526- 4534 [37] (a) Li, X. -Y., Czemuszewicz, R. S., Kincaid, J. R., & Spiro, T. G. J. Am. Chem. Soc. 1989, 111,7012-7023 b) Li, X. -Y., Czemuszewicz, R. S., Kincaid, J. R., Stein, P., & Spiro, T. G]. Phys. Chem. 1990, 94,47-61 [38] Harris, D. C. and Bertolucci, M. D. in Symmetry and Spectroscopy, Dover, 1989, p93-224 [39] Yu, N. -T., Kerr, E. A., Ward, B. and Chang, C. K. Biochemistry, 1983,22, 4534-4540 [40] Nakamoto, K. in Infrared and Raman Spectra of Inorganic and Coordination Compounds, 4th ed., Wiley, 1986, p291-308 [41] Neya, S., Morishima, I. and Yonezawa, T. Biochemistry, 1981, 20, 2610-1614 80 [42] (a) Antonini, E. and Brunori, M. Hemoglobin and Myoglobin in their Reactions with Ligands, Neuberger, A. and Tatum, E. L (Eds. ) North-Holland Publishing Company, Frontiers of Biology Vol.21,1971 (b) James, B. in The Porphyrins, Dolphin,D. (Ed.); Academic: New York, Chapter 6, Vol. V [43] Salehi, A. Ph.D. Dissertation, Michigan State University, 1988, p143-155 [44] (a) Tsutsui, M., Velapoldi, R. A., Hoffman, L., Suzuki, K. and Ferrari, A. J. Am. Chem. Soc. 1969, 91, 3337-3341 (b) Whitten, D. G., Baker, E. W. and Corwin, A. H. J. Org. Chem. 1963, 28, 2363-2368 [45] Kim, D., Su, 0. Y. and Spiro, T. G. Inorg. Chem. 1986, 25, 3988-3993 [46] Shelnutt, J. A., Straub, K D., Rentzepis, P. M., Gouterman, M. and Davidson, E. R. Biochemistry, 1984, 23, 3946-3954 [47] Tait, C. D., Holten, D. and Gouterman, M. J. Am. Chem. Soc. 1984.106, 6653-6659 [48] Wang, M. -Y. R. and Hoffman, B. M. J. Am. Chem. Soc. 1984, 106, 4235- 4240 [49]Spaulding, L. D., Chang,C. C., Yu, N. -T., &Felton, R. H. J. Am. Chem. Soc. 1975 ,,97 2517- 2525 [50] Parthasarathi, N. ,.,Hansen,C Yamaguchi, S. and Spiro,T. G. J. Am. Chem. Soc. 1987,109,3856-3871 [51] Sarrna, Y. A. Spectrochimca Acta, 1989, 45A, 649-652 [52] Prendergast, K. and Spiro, T. G. J. Am. Chem. Soc. 1992, 114, 3793-3801 [53] Felton, R. H., Yu, N.-T., O'Shea, D. C. and Shelnutt, J. A. J. Am. Chem. Soc. 1974, 96, 3675-3676 [54] Hofmann, J. A. and Bocian, D. F. J. Phys. Chem. 1984, 88, 1472-1479 [55] Schcidt, W. R., Summerville, D. A. and Cohen, 1. A. J. Am. Chem. Soc. 1976, 96, 6623-6628 [56] Shelnutt, J. A., Dobry, M. M. and Satterlee, J. D. J. Phys. Chem. 1984,88, 4980-4987 CHAPTER 3 Optical and Vibrational Spectroscopies of Oxovanadyl Porphyrin Compounds, Evidence of Hydrogen Bonding to the Oxo Ligand 3.1 Introduction Heme preteins are involved in a wide range of functions in nature. The ._ oxygen (.02) transporting and storing proteins hemoglobin and myoglobin reversibly bind 02 with an Fe2+ proroporphyrin active site. Iron protoporphyrins are also found at the active sites of catalytic heme proteins. Catalytic heme proreins initially bind 02 with iron in the ferrous state. The reaction cycle, however,involves a sequence of steps that change the oxidation states of both _ the iron and the oxygen The active intermediates of many catalytic heme proreins are either known or postulated to go through oxoferryl complexes. Oxoferryls are known as intermediates in catalases and peroxidases [l]. cytochrome d [2]. cytochrome c oxidase [3-5] and the P-450 enzyme [6]. The reported vibrational frequencies for u(O=FeW) in proteins ranges from ' 745 crn‘1 in lactoperoxidase compound II [7] to 804 cm'1 in cytochrome c oxidase [3-5]. In previously reported oxoferryl model compounds u(O=FeIV) ranged from 807 cm'1 [8] to 841cm'1 [9] in six coordinate complexes while five 81 82 coordinate complexes ranged from 841 cm'il [10] to 852 cm‘1 [ll-12}. The large range of observed vibrational frequencies for 12(O=Few) has been attributed to the strength of the trans ligand [13] and hydrogen bonding to the oxo ligand [14- 15]. The large range of vibrational frequenciesseen in oxoferryl complexes is not seen in vibrational frequencies of Oxy-Feu complexes. Reported frequencies of v(Oz-Fe") ranged from 559 om-l in horse radish peroxidase III [16] to 572 era-1 in myoglobin [17]. In heme model compounds the reported range of frequencies for v(Oz-Fe“) in TpivPP is 562 cm'1 and 573 cm" in protoporphyrin. The reported v(Oz-Fe") frequencies for the heme models is essentially the same as the frequencies for heme proteins. Free oxoiron(IV) porphyrin species. are unstable above -30°C [8-l3]. Therefore, other oxornetals such as chromium [18-20], manganese [21-24], ruthenium [25] and vanadium [26, 33] have been used to gain an understanding of oxoferryl intermediates in the reduction of oxygen. In nature, examples of vanadium are, for the most part, restricted to marine organisms [27-29]. However, the use of vanadium substituted proteins, owing to their stability has proved a useful method in understanding biological functionality [30-31]. To probe environmental effects on oxovanadyl systems 811 et al.[26] have carried out detailed studies of solvent effects on oxovanadyl porphyrin systems. They have shown that a linear correlation occurs between u(V-O) and the solvent acceptor number established by Gutmann [32]. The results they obtained on the oxovanadyl porphyrins were correlated with reported v(Fe-O) of oxoferryl porphyrins. Although hydrogen bonding to the oxo ligand was implied in their work, no direct spectroscopic evidence was given. In this work we show direct eVidence for both intermolecular and intramolecular hydrogen bonding to oxovanadyl porphyrins. Intermolecular hydrogen bonding was verified by the deuterium isotope effect. In the 83 intramolecular hydrogen bonding case, we have established the hydrogen bonding ability of the model compound, naphthyl Kemp's acid porphyrin. Hydrogen bonding from the Kemp's porphyrin is seen in both the acid (NKAP) and amide (NKAmP) forms. The NKAP has a much stronger effect on the oxo- vanadyl vibration which is consistant with the acid form being a better proton donor than the amide form. 3.2 Experimental 3.2.1 OxolS-(8-Kempacid-1-Naphthyl)-2,8,13,17-Tetraethyl-3,7,12,18- Tetramethyl Porphine]vanadium (l) The naphthyl Kemp porphyrin derivatives were all synthesized by Ying Liang. A detailed description of the synthetic process will'be given in her thesis and corresponding publications. Ten mg (0.01 mmol) NKAP, 15ml acetic acid, 450 mg sodium acetate and 100 mg VO(acac)2 (Aldrich, 95%) were placed in a round-bottomed flask fitted with a ground glass joint connected to a reflux condenser. The mixture was refluxed under argon and the progress of the reaction was monitored by UV-vis spectroscopy. When a sample, withdrawn with a pipette, indicated that no more complex was being formed (>95% conversion), the reaction was quenched by adding 15 ml of water. The mixture , was allowed to stand overnight to form the crystal. The product was collected by filtration and washed with water until the filtrate was colorless. The crude compound was purified by running it over a short column of silica gel with methylene chloride as eluant. This gave the pure product which is red in color. Yield: 9.9 mg (92%). FABMS: m/e, 920. UV-vis: Am at 417.5 nm is identified as the porphyrin Soret band The IR spectra show a Strong absorption band at 978 cm'1 .characteristic of an oxovanadyl vibration. Methylene chloride was freshly distilled before use. VO(acac)2 (vanadyl acetyl acetate, 95%) and DBN (1,5-diazabicyclo[4,3,0]one-5-ene, 95%) were 84 purchased from Alderich and used as received. Acetic acid-d (98 atom%D) was purchased from Alderich. HOAc (glacial) was from EM science and NaOAc (anhydrous) were purchased from J. T. Baker. 3.2.2 Instrumentation UV-vis spectral measurements were carried out on a Perkin-Elmer lambda-5 spectrophotometer. Infrared spectra were obtained by layering the porphyrin complex (1) on a NaCl plate, spectra were then recorded on a Nicolet IRI42 spectrometer. FAbMS (fast atom bombardment mass spectra) were recorded on a JEOL HX-llO HF double focusing spectrometer operating in the positive ion detection mode. Resonance Raman spectra were recorded on a computer- controlled Spex 1401 double monochromator equipped with photo-counting electronics. Laser excitation at 406.7nm and 413.1nm was provided by a Coherent-Innova Model 90K krypton ion laser. Excitation at 441.7nm was from a He-Cd laser. Preparation conditions for individual samples is given with each spectrum. The integrity of all RR samples was monitored before and after laser excitation by absorption spectroscopy. 3.3 Results Figure 3.1 shows the structures of naphthyl Kemp's acid porphyrin(NKAP) in (a) the free base and (b) the oxovanadyl forms. The crystal structure of the oxovanadyl amide form shows that the proton of the amide is oriented so that the vanadyl, oxo, and acidic protons are at 180°. This orientation is set for maximum hydrogen'bonding to the axial oxo ligand. The crystal structure of the free base NKAP reveals that a water molecule is oriented into the "pocket” such that hydrogen bonding occurs between the acid proton and the oxygen of water. Correspondingly, a water proron and the carbonyl oxygen are also hydrogen bonded. Figure 3-1 (a) The Structure of free base naphthyl Kemp's acid porphyrin showing the acid proton coordinated over the center of the porphyrin ring. (b) The structure of oxovanadyl napthyl Kemp's acid porphyrin showing the potential for hydrogen bonding between the acid proton and the oxo ligand. 87 3.3.1 Optical Properties Figure 3.2 shows the optical absorption spectra of oxovanadyl naphthyl Kemp's acid porphyrin {NKAP(VO)} in the protonated (solid line) and deprotonated (dashed line) forms. In the deprotonated compared to the protonated acid we observe a 4nm blue shift in the Soret band from 417.5nm to 413.5nm. ‘ This blue shift is accompanied by a 30% increase in the extinction coefficient. If, however, we calculate the oscillator strength from: f=4.33x10‘°Je(13)dt-J eq.l 0 we find that it remains unchanged. Since the Soret band arises from the allowed 1t to It“ transition of the porphyrin system the relatively small shift of the Soret band implies that the hydrogen bond interaction effects are localized to the V0 ‘ portion of the molecule. The Q bands in the visible region of the absorption spectrum show no apparent effect from the removal of the acidic proton. This implies that the energy levels and subsequent orbital interactions are only slightly affected by hydrogen bonding to the axial ligand. Figure 3.3 shows the optical spectra of oxovanadyl octaethyl porphyrin [OEP(VO)} in pure methylene chloride (solid line) and a 1:1 mixture of methylene chloride and acetic acid. The optical spectra of six-coordinated MPs, when compared to five-coordinated MPs, all have characteristic red shifted Soret bands. The bands of OEP(VO) in methylene chloride and acetic acid show no change in the position of the Soret band. Thus, there is no indication of axial coordination by acetic acid to the vanadyl to give a six-coordinate compound. -We also tested the coordinating properties of tetra butyl ammonium acetate. The acetate ion, Figure 3.2 The optical absorption spectra of oxovanadyl naphthyl Kemp's acid porpyrin, (a) (-) in pure methylene chloride and (b) (~---) in methylene chlorided with 1-2% 1,5- diazabicyclo[4,3,0lone-Sene. ABSORBANCE 300 f1) ‘0 1.3.3.3" Iflfiu\flwuesve~wo— L l 400 500 600 WAVELENGTH (nm) UL L 700 Figure 3.3 The optical absorption spectra of oxovanadyl octaethyl porphyrin (—) in pure methylene chlorided and (~---) in a mixture of 50% methylene chlorided and 50% acetic acid. 9i I I I I I \--------- 400 500 600 700 300 wit Th C0 92 with its full negative charge should be a better coordinating ligand than acetic acid. However, we found that the acetate ion does not coordinate to OEP(VO). The optical results are verified by resonance Raman spectra that show no six- coordinate character for OEP(VO) in the presence of acetic acid or the acetate ion. 3.3.2 Resonance Raman Figures 3.4a, and 3.4b show the RR spectra, with kg; = 413.1 nm, of OEP(VO) in CH2C12 with various concentrations of acetic acid (HOAc) or tetra butyl ammonium acetate. Figure 3.4c is the RR spectra of OEP(VO) in DOAc and figure 3.4d is the RR spectrum of OEP(VO) in CH2C12 with tetra butyl ammonium acetate. In spectrum 4a with OEP(VO) in CHzClz and 5% HOAc the OEP carbon-nitrogen skeletal modes, are unaffected by the addition of HOAc. Previous studies of oxovanadyl porphyrins have established that the V0 stretching frequency is in the 1000 cm‘1 region, thus, we can assign the peak at 991 cm‘1 to stretching of the V0 bond. In addition to the V0 vibrational mode at 991 cm'1 the appearance of an additional band at 974 cm'1 is detected. The intensity of this band is dependent on the concentration of acetic acid present in the sample. This intensity dependence is evident in spectrum 3.4b where we have OEP(VO) in 75% acetic acid and 25% methylene chloride solvent. In spectrum 3.4c we used deuterated acetic acid at a 1:1 ratio with methylene chloride. Spectrum 3.4c is the first reported evidence that the oxovanadyl bond is sensitive to direct. hydrogen bonding. Hydrogen bonding is a factor in the acceptor number calculated by Gutmann [31]. However, there are no previousely reported cases of a deuterium isotope effect to verify the presence of hydrogen bonded (VO)porphyrins. The three wavenumber shift to lower energy for u(VO) is consistent with an increased mass for the (H---O)-V oscillator system. To verify Figure 3.4 Resonance Raman spectra obtained from 413.1 nm excition of OEP(VO) in (a) 95% methylene chloride and 5% acedic acid (b) 25% methylene chloride and 75% acedic acid(c) in 50% methylene chloride and 50% deuterated acetic acid (d) in methylene chloride and an excess of tetrabutyl ammonium acetate. Optical absorption spectra were taken before and after the Raman spectra to ensure the integrity of the. sample. ‘12 i? 1 O 2'; a) 5% HOAc/CH2C12 " 1 ° .. I O h I" " " 9 E’- ..n " f V Ll “\K‘W w\‘~_—/\/\\J\_~_ 1 3.4. 8 I b) 70% HOAchIizClz ' r: S U) I“ C 0| .2 c)50%DOAc/CH2C12 C \‘N \ . , d) excess (t-BuMNOAc 1 0100 1200 1 400 1 600 Roman shift (cm-l) 95 that this is indeed hydrogen bonding and not a six-coordinate complex we used an excess of tetra butyl ammonium acetate. In spectrum 4d we can see that v(VO) is located at 991 cm'1 or in the same position as 0(VO) in pure CH2C12.. According to Gutmann's calculations acetic acid has a high coordination number. From our results, however, we conclude that the major effect of acidic acid on OEP(VO) is the formation of a hydrogen bond between oxovanadyl and the acid proton. Figure 3.5b is the RR spectrum of NKAP(VO) in pure methylene chloride. Using the assignments of Abe et. al. [34] we can establish that the vibrational modes between 1000 and 1700 cm“1 are all identified as etio porphyrin . vibrations. The vibrational mode at 978 cm‘1 we assign as. the V0 stretch of the intramolecular hydrogen bonded NKAP(VO) complex. Evidence that the vibrational mode at 978 cm‘1 is due to intramolecular hydrogen bonded oxovanadyl comes from Figure 3.5a. Figure 3.5a is the RR spectrum of NKAP(VO) in methylene chloride and 1-2% DBN. Although the porphyrin core vibrations are unaffected by the addition of DBN, the peak at 978 cm"1 is shifted 13 cm"1 to higher energy. This shows an increase in the V0 bond strength. We interpret this as a removal of the acidic proton and the corresponding hydrogen bond. Removal of the hydrogen bond allows the electron density of the hydrogen bond to localize into the V0 bond. Figure 3.6a is the RR spectrum of OEP(VO) in pure methylene chloride. We can identify the usual porphyrin core vibrations; the additional peak at 991 cm'1 is identified as the V0 stretching vibration. Figure 3.6b is the RR spectrum of OEPfVO) in methylene chloride and excess DBN. We detect no signs of peak shifts from Fig. 3.6a and conclude that DBN, at the concentrations that we are using, does not act as a sixth ligand to the vanadyl. Figure 3.5 Resonance Raman spectra obtained from the 413.1 nm , excitation of oxovanadyl naphthyl Kemp's acid porpyrin, (a) in pure methylene chloride and (b) in methylene chlorided with 1-2% l,5-diazabicyclo[4.3,0lone-S-ene. Optical absorption speCtra were taken before and after the Raman speCtra to ensure the integrity of the sample. RELATIVE INTENSITY 97 991 [(13 a. NKAP(VO)/CH2C12/DBN 413.1nm@10mW 07 b. NKAP(VO)/CH2C12 413.1nm@10mW {5 I“) O7 '2 1.0 CO LC 4 .. I\ . l a e N ‘0 "' to w 1'3 a 9:: 1 c ,7, l '— N N “5 N o—- a '- 1‘ w i \ A l a 10‘ E «e b 1 i "' ”f3 E J \ j \A U \W 1000 12:00 1.400 _ 1600 RAMAN SHIFT (CM—1) Figure 3.6 Resonance Raman spectra obtained from 413.1 nm excition of OEP(VO) in (a) in pure methylene chloride and(b) in methylene chlorided with an excess of 1,5-diazabicycio [4,3,0]one-5-ene. Optical absorption spectra were taken before and after the Raman spectra to ensure the integrity of the sample. 99 i 3 f :71 I A g '2 I . ., t a) 1n CH2C12 w/out DBN l a i 3 E .- E A ,__'_:::—> Intensity b) in CH2C12 w/ DBN ”/1 kw/th/ VJM 1159 1223 . 1260 1514 I41 142% C) ‘7‘»--.- “.---.— —. .-——— - .- ' ._ .-..—oco -’-o~-- 1000 1200 1490 ‘500 RomGnSMHikfihd) 100 Figure 3.7 is the RR spectra of oxovanadyl naphthyl Kemp's amide- porphyrin (NKAmP(VO)) in methylene chloride (a) with DBN and (b) without DBN. DBN is a relatively weak base compared to amide and we expect the amide to remain fully protonated in the presence of DBN. Comparison of 3.7a with 3.7b shows that no effect from DBN is detectable. We established above that DBN does not coordinate to the vanadyl. Since DBN does not coordinate and it is a relatively weak base compared to amide we can predict that the frequency of V0 will not be affected in the presence of DBN. The results seen in Figs. 3.7a and 3.7b verify that no effect from DBN is seen. Figure 3.8 is the RR spectra of onp(v0) and NKAmP(VO) N-methyl Imidazole at either 406.7 nm or 441.1 nm laser excitation. At these wavelengths we can selectively excite the six-coordinate complex (441.1 nm) or a mixture of the five and six-coordinate complexes (406.7 nm). In both the-five and six- . coordinated NKAmP(VO) forms we see a3 cm'l shift of v(VO) relative to the u(VO) of OEP(VO). From these results it is evident that the effect on v(VO) can not be accounted for simply in terms of solvent to metal interactions. The properties of N-methyl imidazole are similar to those of pyridine and we expect similar interactions from these solvents. The V0 vibration in the five-coordinate OEP(VO) complex is at 989 cm'1 and at 957 cm'1 for the six-coordinate complex. The V0 vibration of NKAmP(VO) is 988 cm-1 for the five-coordinate compound and 954 cm'1 for the six-coordinate compound. The six-coordinate compound for both NKAmP(VO) and OEP(VO) have the same 3 cm'1 difference for 0(VO) in N-methyl imidazole as is scen in methylene chloride. In the five-coordinate ' compounds, in N-methyl imidazole, ut V0) for NKAmP(VO) is unaffected relative to methylene chloride while v(VO) for OEP( V0) is down shifted 2 cm'1 relative to u(VO) in methylene chloride. This implies that the solvent effects on the five- coordinate compounds that changes mm). are oriented to the oxo side of the Figure 3.7 Ni Resonance Raman spectra obtained from the 413.1 nm excitation of oxovanadyl naphthyl Kemp's amide porpyrin, (a) in pure methylene chloride and (b) in methylene chlorided with l,5-diazabicyclo[4,3,0]one-5-ene. Optical absorption spectra were taken before and after the Raman spectra to ensure the integrity of the sample. SITY RELATIVE INTEN 102 ”‘_ ........__.-, b Q) G) O) ”- \ Q Ma) \VL a. NKAmP(VO)/CH2C12/DBN 413.1nm@10mW b. NKAmP(VO)/CH2C12 413.1nm@101nw 8 V 01 n ‘1" to "" “n <10 c: N e R? I?) m .0 "' (O N in -- sr {\ "" 1 000 I 1200 14b0 1600 RAMAN SHIFT (CM—1) 103 Figure 3.8 (a) Resonance Raman spectra obtained from the 406.7 nm excitation of oxovanadyl naphthyl Kemp's amide porpyrin, in N-Methyl Imidizoie. (b) Resonance Raman spectra obtained from the 406.7 nm excitation of oxovanadyl octaethy porpyrin, in N-Methyl Imidizole. (c) Resonance Raman spectra obtained from the 441.8 nm excitation of oxovanadyl naphthyl Kemp's amide porpyrin, in N-Methyl Imidizole. (d) Resonance Raman spectra obtained from the 441.8 nm excitation of oxovanadyl octaethy porpyrin, in N-Methyl Imidizole. 104 954.5 95 989 ’ 9! l\ O o {I O o . 'T 990 1138.4 E to Z Lt-L-J 0 Z LtJ 2 5... S LLJ O: b C d 900 1000 1 100 RAMAN SHIFT (CM- 1) 1200 105 porphyrin. That is to say, the vanadyl side is unrestricted and if solvent effects are oriented solely to the vanadyl, then solvents should effect both NKAmP(VO) and OEP(VO) vibrations equivalently. 3.4 Conclusions We have presented proof for both intramolecular and intermolecular hydrogen bonding effects on oxovanadyl porphyrin systems. In the intermolecular case we have presented the first deuterium isotope effect on the vanadyl-oxo vibration. The shift to a lower energy when DOAc is used in place of HOAc has the predicted effect since there is an effective mass increase in (D---O)-VOEP versus (H--°O)-VOEP oscillators. Results from Yan and Czemuszewicz [35] working with vanadyl protoporphyrin IX and different alcohols as solvents they report a deuterium isotope effect that was previously undetected [26]. The deuterium isotope effect that they see has the v(V =0) vibration shifting to a higher energy than the proton effect. To explain this difference we need to consider the proton releasing ability of the donor solvent The PKa of acidic acid is 0.7 while the Pka of methanol is 17 [36]. The H-OAc bond is considerably weaker than the H-OCH3 bond. The effect that this has is that the hydrogen bond of V=O---H-OAc is considerable stronger than the hydrogen bond of V=O---H-OAc in other words the V=OmH of the acid is more of an individual unit, therefor, more prone to the effects of the increased mass of the deuterium substitution. We have demonstrated the intramolecular hydrogen bonding ability of the recently synthesized Naphthoic Kemp's acid porphyrin and its derivatives. The hydrogen bonding ability was clearly demonstrated by removal of the acidic proton with the demonstrated non-coordinating base DBN. The resulting effects 106 were manifested in the oxo-vanadyl vibration. The likelihood that steric effects, not hydrogen bonding, is the cause of this interaction seems highly unlikely since the amide form would be equally, if not more, hindering than the acid form of naphthoic Kemp's porphyrin. Furthermore, we rule out the possibility of purely ionic interactions, since removal of the acidic proton of NKAP causes v(O-V) to shift back to the same frequency as v(O-VOEP). The addition of DBN to NKAmP has no effect on v(O-V), this implies that some factor is still present to influence v(O-V) and we conclude this factor to be hydrogen bonding to the oxo ligand. 107 REFERENCES [1] Hewson, W0. and Hager, LP. in The Porphyrins; Dolphin, D.;Ed.; Academic: New York, 1979, Vol. VII, Chapter 6 [2] Loehr, T. Biochem. 1991, 30, 11485-11489 [3] (a) Varotsis, C. and Babcock, G.T. Biochem 1990, 29, 7357-7362 (b) Varotsis, C.; Woodruff, W.H. and Babcock, G.T. J. Biol. Chem 1990, 265, 1 1131- (c) Varotsis, C.; Woodniff, W.H. and Babcock, G.T. J. Am. Chem. Soc. 1989, l l l, 6439- (d) Varotsis, C.; Zhang, Y.; Appelman, EH. and Babcock, G.T. Proc. Natl. Acad. Sci.,U.SA. 1992, 90, 237-241 [4] (a) Han, S.; Ching, Y.-c. and Rousseau, D.L. Biochemistry, 1990, 29, 1380- (b) Han, S.; Ching, Y.-c. and Rousseau, D.L. Proc. Natl. Acad. Sei.,U.S.A. 1990,87, 2491- (c) Han, 8.; Ching, Y.-c. and Rousseau, D.L. Proc. Natl. Acad. Sci.,U.S.A. 1990, 87. 8404— [5] (a) Ogura, T.; Takahashi, S.; Shinzawa-ltoh, K; Yoshikawa, S. and Kitagawa, J. J. Biol. Chem. 1990, 265, 14721-14723 (b) Ogura, T.; Takahashi, 8.; Hirota, S; Shinzawa—Itoh, K; Yoshikawa, S.; Appelman, EH. and Kitagawa, J. J. Am. Chem. Soc. 1993. 1 15, 8527-8536 [6] Watanabe, Y. and Groves, J.T. in The Enzymes, Chap. 9, Academic Press, Inc., New York, 1992 [7] Reczek, C.M.; Sitter, AJ. and Terner, J. J. Mol. Spec, 1985?6?, [8] Schappacher, M.; Chottard, G. and Weiss, R. J. Chem. Soc., Chem. Commun., 1986, 93-94 [9] Gold, A.; Jayaraj, K.; Doppelt, P.; Weiss, K.; Chottard, 6.; Bill, 15.; Ding, X and Trautwein,A.X. J. Am. Chem. Soc, 1988, 110, 5756-5761 [10] Czemuszewicz, R.S. and Macor, K.A. Ramon spectrosc., 1988, 19, 553-557 [1 l] Bajdor, K. and Nakamoto, K. J. Am. Chem. Soc, 1984, 106, 30456046 1m [12] Ptoniewicz, Z.L.; Bajdor, K. and Nakamoto, K J. Phys. Chem, 1986,90, 1760-1766 [13] Oertling, W.A.; Kean, R.T.;Wever,R. and Babcock,G.T. Inorg. Chem, 1989, 29, 2633-2645 [14] Sitter, AJ.Reczek, CM. and Temer, J. J. Biol. Chem, 1985, 260, 7515-7522 [15] Hashimoto, S.: Tatsuno, Y. and Kitagawa, '1‘. Proc. Natl. Acad. Sci., 1986,83, 2417-2421 [16] VanWart, HE. and Zimmer, J. J. Biol. Chem., 1985, 260, 8372-8377 [17] Tsubaki, M.; Nagai, K. and Kitagawa, T. Biochemistry, 1980, 19, 379-385 [18] Groves, 11‘. and Kruper,WJ.Jr. J. Am. Chem. Soc. 1979,101,3413-3415 [19] Nill, K.H.;Wasgestian, F. and Pfeil, A. Inorg. Chem, 1979, 18, 564-567 [20] Groves, J.T. and Kruper,WJ.Jr. and Nemo,T.E. J. Mol. Catal. 1980.7,169- 177 [21] Tabushi, I. and Koga, N. J. Am. Chem. Soc. 1979, 101, 6456—6458 [22] Hill, CL and Schardt, B.C. J. Am. Chem. Soc. 1980, 102, 6374-6375 [23] Groves, J.T. and Kruper, WJJr. and Haushalter, R.C. J. Am. Chem. Soc. 1980, 102, 6375-6377 [24] Willner, 1.; Otvos,J.W. and Calvin, MJ. J. Chem. Soc. Chem Commun. 1980, 964-965 [25] Groves, J.T. and Ahn, K.-H. Inorg. Chem. 1987, 26, 3833-3835 [26] Su, O.Y.; Czemuscewicz, R.S.; Miller, LA. and Spiro, T.G. J. Am. Chem. Soc. 1988, 110, 4150-4157 [27] Vilter, H. Phytochemistry 1984, 23, 1387-1390 [28] Carlson, R.M.K. Proc. Nat. Acad. Sci. USA 1975, 72, 2217-2221 [29] Carlisle, D.B. Proc. Roy. Soc. London Ser. B. 1968, 171,31 [30] Buchler, J.W.; Puppe, L.; Rohbock, K. and Schneehage, H.H. Ann. N. Y. Acad. Sci. 1973, 206, 116— [31] Pettersen, R.C. and Alexander, LE. J. Am. Chem. Soc. 1968, 90, 3873-3875 109 [32] (a) Mayer, U.; Gutmann, V. and Gerger, W. Monatash. Chem, 1975.106, 1235-1257 (b) Gutmann, V. Electrochem. Acta., 1976, 21, 661-670 (c) Gutmann, V. The donor-Acceptor Approach to Molecular Interactions, Plenum: New York, 1978 5323]6Lin, M.; Lee, M.; Yue, KT. and Manilli, L.G. Inorg. Chem. 1993, 32, 3217- 2 [34] Abe, M.; Kitagawa, T. and Kyogoku, Y. J. Chem. Phys. 1979, 69, 4526-4534 [35] Yan, Qing and Crenuszewicz, Roman S. Presented at the International Conference on Raman Spectroscopy in Hong Kong [36] March, Jerry Advanced Organic Chemistry John Wiley & Sons, Inc.: New York 3rd Ed. 1995 CHAPTER 4 Cyanide Titrations of (Cl' )Fegmporphyrins Followed by Optical Spectrosopy 4.1 Introduction 4.1.1 Sol-gel systems Sol-gel processes involve the mixing of two liquids that react to give a uniformly mixed material. The concept of mixing liquids allows for complete interactions at the molecular level. If the liquids, once mixed, undergo a chemical ‘ reaction, a polymer-like material results. The reaction may occur by itself or it may, in fact, require a catalyst. The reaction proceeds with an increase in viscosity of the liquids; when it reaches a solid state it is known as a solid gel or sol-gel. Ninety-five percent of the sol-gel processes being used today involve inorganic oxides. To explain this process an example, such as, barium-oxygen and titanium-oxygen will be helpfull. In the barium-titanate crystal, barium is next to an oxygen that is next to a titanium, and so on. This leads to a very ordered array within the crystal structure. However, in the sol-gel process, polymerization of these compounds are, indeed, crystalline like in the short range; but, due to the continuous molecular exchange that occurs in liquids the overall product is amorphous and very porous. Trapped within the pores are the solvents that were used; indeed, the sol-gel product, in this respect, resembles a sponge. The product 110 111 or sponges, if you will, may be treated with acid, base or whatever appropriate agent will preferentially remove the trapped solvent leaving behind a porous oxide that is somewhat disordered. If you heat the material to evaporate the solvent you end up with a very dense porous material. The more you heat the compound the denser it gets. This density factor comes about from the large surface area created by the pores. In the heating process the system reduces the amount of free energy present and it does so by collapsing in on itself, thus, becoming denser and denser. Now we have a method of making dense oxide materials at relatively low temperatures. In the case of the gel the viscosity increases as the gel starts to form. At a certain point in the gel's formation, the viscosity is expected to be such to allow for molding of the material into specific shapes such as drawn fibers, shaped blocks, thin films and many more [1]. 4.1.2 Cyanide detection systems If we consider the sponge like nature of sol-gel materials, in the dried form we have a product with interconnecting pores that are approximately 100 Angstroms in diameter. With controlled heating, the pore size can be adjusted to smaller sizes. The pores may be impregnated with organic molecules with any of a variety of properties. In our case, we were looking for a suitable material to act as a cyanide detector. The optical properties of iron porphyrins are well established [2]. Likewise the strong propensity of cyanide anions for ferric porphyrins is well known [3]. It then seems reasonable to make thin film sol-gel sheets and impregnate these with a ferric porphyrin to act as a cyanide ligand trapping agent. The optical 112 properties of the unligated and the ligated forms could be used to establish the concentration of cyanide present in the test solution. 4.1.3 Properties of substituted phenyl tetraphenyl porphyrins The spin state of an Femporphyrin is dependent on the ligand field strength of the equatorial [4-7] and axial [8-10] ligands. The more electron withdrawing the macrocycle is, the higher the spin multiplicity[1l]. Thus, it is possible to control the spin state by varying the electron withdrawing nature of the peripheral substituents around the porphyrin macrocycle. To understand the environmental influences, various model complexes have been synthesized and studied. Substituents on the phenyl ring of tetraphenyl porphyrin (TPP) significantly alter the redox potential and subsequent binding properties of the metal center. The redox potentials of o- bonded perfluoroarylFemporphyrins and alkylFemporphyrins have been reported by Kadish and coworkers [ll-15]. The work of Kadish et al. showed that the redox potentials for Femtetraphenyl porphyrins was significantly affected by axial coordinated ligands with differing electron withdrawing properties. The more electron withdrawing the axial ligand is the the higher the oxidation potential of the iron porphyrin. 4.2 Experimental All samples were prepared in freshely distilled acetonitrile as solvent. Tetrabutyl ammonium cyanide was used as purchased from Fluka chemicals. Stock solutions were freshly prepared at the beginning of each series of titration's. Titrations were followed by electron absorption spectroscopy on a Cary 214 spectrophotometer. A one centimeter path length quartz curette fitted 113 with an extended neck was used as the titration vessel. Tetrabutyl ammonium cyanide samples were prepared by weighing out tetrabutyl ammonium cyanide and dissolving in a known volume of freshly distilled CH3CN. 4.3 Results Analysis of the coordination of cyanide to the different porphyrins used in this study can be made though the following equations: Keg (Cl')Fe"P+ CN ¢PCN Eq.l K... (Cl’)Fe' PCN+ CN‘ ¢>P(CN)2 Eq. 2 Keq (CF)Fe'P+2CN‘¢>P(CN)2 Eq. 3 In all the titration curves of the various porphyrins there is only one observable isosbestic point. This leads to the conclusion that only one of the equations above is valid for our studies. Equation 2 can be ruled out as the binding scheme since it has two stepwise additions of cyanide to the porphyrins. Equation 3 is unlikely since his coordinated cyanide porphyrins require significantly higher concentrations of cyanide than those with which we are working with [17]. The likely conclusion then, is that eq. 1 accounts for the binding scheme of cyanide to the various fluoro-substituted phenyl tetraphenyl porphyrins. From eq. 1 we define Keg as: [NCFemP] K" "(Fe"PrC~1 114 Experimental binding data are often analyzed with the fractional saturation term 7 [3a], so, l" ...+ 007 ‘p ‘ -- 8)»).-- J, 0.5 “- 0.50- 0,34b WAVELENGTH (nm) Figure 4.3 Titration curve of (pentafluoro) tetraphenyl porphyrin (Fem)chloride (FeIn (PentaF)TPPCl) in acetonitrile (CH3CN) titrated with tetrabutyl ammonium cyanide in CH3CN, 1.0a 0.9 m ABSORBANCE 119 ,,- WAVELENGTH (nm) Figure 4.4 Titration curve of Octaethyl porphyrin (Fem)chloride (FemOEPCl) in acetonitrile (CH3CN) titrated with tetrabutyl ammonium cyanide in CH3CN. 120 l ..- —rrfl))-~ u 1.04 } f 0.90- E!“ 0.0-- 0.7- l 0.0 r 0.54 ABSORBANCE 0.4 a 0.3'1 //////Z/’ /’ 0.24 /’/ // U 0.1 4 L I zoo 450 500 WAVELENGTH (nm) --\ 000 Figure 4.5 Titration curve of tetraphenyl porphyrin (Fem)chloride (FemTPPCl) in acetonitrile (CH3CN) titrated with tetrabutyl ammonium cyanide in CH3CN. m.) 1 0.9 <1 0.0": 0.7 0 0.50 ABSORBANCE 0.4 J» 0.3 <1 0.2" 0.1 ,., zoo 450 soo soo WAVELENGTH (nm) Figure 4.6 Titration curve of porphycene (Fem)chloride (FemPor) in acetonitrile (CH3CN)titrated with tetrabutyl ammonium cyanide in CH3CN. 122 4.4 Discussion The binding constants of cyanide to the electron withdrawing properties of the peripheral substituents show a direct correlation. The binding constants increase with an increase in electron withdrawing ability of the substituents. This correlation holds true with the exception of porphycene, which we will address later. Table 2 is a summary of redox potentials, measured by Kadish and coworkers, for o-bonded perfluoroarylFemporphyrins or o-bonded chloroFemporphyrins. The properties of these compounds show a direct correlation with the electron withdrawing ability of the axial ligand. First, we can see that the axial ligand is significant in determining the spin state of the metal. If the ligand is more electron withdrawing (better rt-acid) then the core iron is more likely to be high spin. Second, it is apparent that the axial ligand is highly significant factor in the redox potential in that, the more electron withdrawing the axial ligand is, the higher the oxidation potential. Third, and more significant to the present work, is that the peripheral substituents of the ring are also a major factor in the redox potential of the iron core. The peripheral substituents, like the axial ligand, increase the oxidation potential when they are electron withdrawing. In our binding studies the cyanide anion is strongly nucleophylic so it would be intuitive that the higher the oxidation potential of the iron, the easier the anion would be to bind. If we look at our binding constants in table 1 this is exactly the trend we see. The binding constant for FemOEP-Cl, which has electron donating peripheral substituents, is almost an order of magnitude less than the binding constant for FemT(pentaFP)P-Cl, which is highly electron withdrawing. Table 2 . Half-wave 123 PhCN (0.1) M (T BA)PF5). Reproduced from reference 1 1. Porphyrin (P) axial ligand spin state r.t. OEP TPP (m-Me)-TPP (p-Me)-TPP (CN)4TPP (p-EtZN)-TPP CI C6F5 C6F4H C6H5 Cl C6F5 C6F4H C6H5 Cl C6F5 C6F4H Cl C6F5 C6F4H Cl C6H5 Cl C6H5 hs hs hs I s hs hs hs Is 115 hs hs hs hs hs hs I s hs I s Oxidation 2nd 1st 1.08 1.18 0.87 1.14 0.79 1.3 0.48 1.14 1.38 0.94 1.32 0.86 1.43 0.61 1.57 1.18 1.36 0.92 1.31 0.84 1.56 1.13 1.33 0.89 1.28 0.82 1.42 1.03 0.73 0.59 0.58 0.47 1st -0.54 -0.59 -0.64 -0.93 -O.39 -0.42 -0.45 -0.7 -0.33 -0.42 -0.46 -0.34 -0.45 -0.49 0.16 -0.03 -0.48 -0.83 reduction 2nd -1.26 -1.3 -1.28 -1.09 -1.06 -1.06 -1.08 -1.08 -1.06 -1.09 -1.09 -1.07 -0.43 -0.96 -1.17 potentials of different (P)Fe(R) and (P)Fe(Cl) complexes in 3rd -1.95 -1.71 -1.72 -1.71 -1.73 -1.76 -1.74 -1.77 -1.77 -1.74 -0.85 In the iron porphycene (Fig. 4.7) we have a molecule that apparently does not follow the trend of the porphyrins in our study. The 8 and meso carbons of the porphycene molecule are all covalently bound to a hydrogen. Hydrogen would conceptually place porphycene somewhere between Fem'l'(PMP)P-C1 and FemTPP-Cl in our cyanide binding study. However, the core of porphycene is larger than the core of regular porphyrins. In all of the porphyrins and the porphycene the fifth ligand is a chloide ion. We know from previous studies that five coordinated metalloporphyrin complexes have the metal out of the plane formed by the four pyrrole nitrogens. The further the metal is pulled out-of-the plane the harder it is to bind a sixth ligand. A possible explanation for the cyanide binding difference of the porphycene, compared to the porphyrins, is that 124 the iron of the porphycene is not pulled out of the plane as far as the iron of the porphyrins. The sol-gel systems discussed in the introduction lend themselves well to the introduction of porphyrins, in particular tetraphenyl porphyrins. Keeping in mind that the emphasis of this study was to find a suitable system to prepare cyanide detectors the porphyrin choosen for this purpose was PFPP. The detection system developed from the implimentation of PFPP into a sol-gel derived titanium carboxylate thin film was found to be stable over a long period of time, with the properties nescessary for acurate detection of cynaide ions over a wide range of concentrations [17]. REFERENCES [1] Mackenzie, J.D. 0E Reports, 1990 [2] Gouterman, M. The Porphyrins, Dolphin, D. ,Ed., Academic: New York Vol. 111 Chapter 1 [3] (a) Antonini, E and Brunori, M. Hemoglobin and Myoglobin in their Reactions with Ligands, Neuberger, A. and Tatum, E L (Eds) North-Holland Publishing CBrE‘pany, Frontiers of Biology Vol. 21, 1971 (b) James, B. in The Porphyrins, phin,D. (Ed.); Academic: New York, Chapter 6, Vol. V [4] Scheidt, W.R. and Reed, C.A. Chem. Rev., 1981,81, 543 [5] Scheidt, W.R.; Geiger, D.K. and Haller, K]. J. Am. Chem. Soc., 1982, 104, 495 [6] Scheidt, W.R.; Geiger, D.K.; Hayes, R.G. and Lang, G. J. Am. Chem. Soc., 1983. 105, 2625 [7] Behere, D.V.; Birdy, R. and Mitra, S. Inorg. Chem, 1984, 23, 1978 [8] Mitra, S. In Iron Porphyrins, Lever, A.B.P. and Gray, H.B., Eds; Addison- Wesley: Reading, MA, 1983, Part 11, Chapter 1 [9] Scheidt,W.R.;Lee,J.Y. Geiger, D.K.;Taylor, K. and Hantano, K. J. Am. Chem. Soc., 1982, 104, 3367 [10] Geiger, D.K. and Scheidt, W.R Inorg. Chem.,1984, 23, 1970 [11] Guilard, R.; Boisselier-Cocolio, B; Tabard, A. Cocolios, P.; Simonet, B. and Kadish, K.M. Inorg. Chem, 1985, 24, 2509-2520 [12] Lancon, D.; Cocolios, P.; Guilard, R. and Kadish, K.M. J. Am. Chem. Soc., 1984, 106, 4472 126 [13] Lancon, D.; Cocolios, P.; Guilard, R. and Kadish, K.M. Organometallics, 1984. 3. l 164 [14] Kadish, K.M.; Tabard, A.; Lee, W.; Liu, Y.H.; Ratti, C. and Guilard, R. Inorg. Chem, 1991. 30, 1542-1549 [15] Kadish, K.M.; D'Souza, F.; Caemelbecke, EV.; Villard, A.; Lee, J.-D.; Tabard, A.; and Guilard, R. Inorg. Chem, 1993, 32, 4179-4185 [16] Neya, S., Morishima, l. and Yonezawa, T. Biochemistry, 1981, 20, 2610-1614 [17] Bunuwilla, D.D.; Torgerson, B.A.; Chang, GK. and Berglund, K. A. Anal. Chem. 1994, 66, 2739-2744 CHAPTER 5 Resonance Raman Evidence of the Hydrogen Bonding Effect on v(M-CO) where M = Fe(II) or Ru(II). 5.1 Background Hemoglobin and myoglobin work in tandem to transport and store oxygen in vertebrates. Hemoglobin is found in high concentrations in red blood cells and myoglobin is found in aerobic muscle tissue. Hemoglobin is an (1282 heterotetramer that cooperatively binds 02 in areas of high oxygen concentration, such as the lungs. The cooperative binding properties allow hemoglobin to become saturated with 02. The 02 saturated hemoglobin is transported, in the blood stream, to areas of lower oxygen concentration, such as respiring tissue, where the 02 is released and delivered to myoglobin. The structural and functional makeup of the individual a and [3 subunits are comparable to myoglobin [1]. Myoglobin, a single ploypeptide, is a small globular, oxygen storing protein, made up of eight a-helices (known as A-H) that form a pocket that constitutes the active site of the protein. The active site of these proteins contains an iron protoporphyrin IX with iron in the reduced (Fe(Il)) oxidation state [2-5]. The iron atom is coordinated to the protein moiety through a histidine at position 8 along the F helix. The 127 128 porphyrin through the four pyrrole nitrogens. The remaining binding site of the iron is known as the distal site and this is where exogenous ligand coordination to the iron takes place. In the reduced state iron, in addition to binding oxygen, is subject to binding by a variety of ligands such as CO, NO, nitroso compounds and alkyl isocyanides [6,7]. Although NO is an important ligand in itself, and the nitroso compounds and alkyl isocyanides are pollutants that we are subject to on an increasing basis, CO is the major competing ligand of 02 for binding to the reduced iron. Unhindered model heme compounds are known to favor CO binding over 02 binding by as much as five orders of magnitude [8—13]. However, when the heme is contained within the protein structure, the CO binding affinity is dropped to less than 200 fold over the 02 binding. Thus, the proteins are extremely efficient at ligand discrimination, which translates into preservation of the life form since CO is an endogenously produced molecule [14,15]. For hemoglobin and myoglobin to function efficiently as oxygen transport and storage proteins, they must have appropriate kinetic and thermodynamic properties with respect to oxygen binding. The affinity for oxygen, by hemoglobin, must be on the order of P1 ,2 = 20 to 50 [J.M which is sufficiently low enough to release oxygen in the muscle tissue. In addition to the binding affinity, the association ( k' 0.) and dissociation ( k0,) rate constants must be on the order of k' 0: >106 M-ls-l and k0: >10 s-l so that rapid release and uptake may occur in muscle tissue and capillary beds [16,17]. The proteins must also be slow to autooxidize to the ferric state and be resistant to denaturing. 5.1.1 Proposed ligand discriminations 129 In 1970 Perutz proposed that the oxygen binding energy differences in various hemoglobins and myoglobins was due to steric hindrance induced by the proximal histidine [18]. The reasoning was that the proximal histidine formed a five-coordinate complex that forced the iron out of the plane of the porphyrin ring. The "pocket" left on the distal side, where ligand binding occurs, makes it difficult for oxygen to bind. The more out plane the iron is, the more difficult it is to bind oxygen. This explanation for the differences in oxygen binding energy works for the case of oxygen but does not explain the protein's ability to discriminate between different ligands. The question of distal effects as ligand binding mediators in heme proteins has resulted in efforts to discern the role of the E7 distal histidine of the primary amino acid sequence [19-30]. In hemoglobin the E7 histidine is situated closest to the ligand binding site [19,24]. Two possible effects of histidine E7 have been used to explain the phenomenon of ligand discrimination in hemoglobin and myoglobin. One school of thought is that the primary effect of histidine E7 is that it exerts its effect through steric hindrance [14,31] and the other school of thought invokes the hydrogen bonding effect. The argument for the steric hindrance effect arises from the fact that CO is at its maximum binding efficiency when it can bind end on in a linear conformation of Fe-C-O with CO perpendicular to the porphyrin plane. In this orientation the orbital overlap, in the n-back-bonding model of transition metal carbonyl complexes, is at a maximum. Oxygen, in contrast to CO, binds end on in a bent conformation. The bent orientation of Fe-02 would not be subject to steric effects of E7 since its hinderance effects would decrease . The crystal structure of the CO-pyridine-tetraphenylporphyrin complex shows that the CO and pyridine ligands are bound to the metal center in a linear orientation [32]. Recently, however, Gerothanassis et al. [33] have suggested 130 that the Fe-CO binding orientation of an unhindered "picket fence" model is actually bent. The basis for the bent assignment was the calculated 13C shielding tensor, which is sensitive to the Fe-CO binding geometry. The calculations were carried out by using data from 13C CP MAS NMR spectroscopy. X-ray and neutron diffraction studies of CO bound hemoglobin and myoglobin show various orientations of the Fe-CO system; some are bent, others are tilted, while others are bent and tilted [34-39]. This variety of Fe-CO binding orientations would imply that hindrance is not the only factor that can effect the ligand binding discrimination in heme proteins. Kim and Ibers have recently presented crystallographic data on a CO bound "capped" porphyrin. Their study showed that the Fe-C-O bond angle was as much as 7° from linearity, yet there was no detectable effect on the CO binding affinity [40]. Kim et al. have shown results with model compounds that have reduced CO affinity while retaining a linear configuration [41]. 5.2 Mutagenisis of HisE7 and the effect on C0 affinity Site directed mutagenesis studies of histidine E7 have given no compelling evidence that steric hindrance is the key factor in ligand binding discrimination. The argument for steric hindrance centers on the concept that larger residues would have a greater hindrance while smaller residues would have a lessened hindrance effect. However, the results to date have not shown these effects. The HisE7Leu mutant has a 30-fold increase in CO affinity over 02 affinity [42-44]. This is an exceptionally "high" CO affinity since the norm for other E7 mutants is 2 to 5-fold. In the Gly, Ala and Val mutants, which are considerably smaller side chains than Leu, the CO affinity is at most 5-fold over that of 02. The HisE7Tyr mutant effectively decreases CO affinity by 6-fold [45]. It has been shown that in the mutation HisE7Tyr, in metmyoglobin, that the tyrosine side chain chelates 131 directly to the ferric iron atom [45,46]. Table 1 is a summary of the reported CO and 02 affinities of HisE7 mutants. From the mutagenesis studies presented above, it would appear that steric hindrance is not the major factor of reduced CO affinity in myoglobin. It would seem most likely that the remaining determining factors are the need to displace distal water molecules and/or direct unfavorable electrostatic effects between the ligand and the adjacent amino acid residues. It has been suggested by Rohlfs et al. [47] that a major kinetic barrier to 02 and CO binding involves the disruption of hydrogen bonding between HisE7 and an adjacent water molecule found in approximately 80 percent of the crystal structures of native or wild type deoxymyoglobins [43]. It is proposed by Quillin et al. that mutation of HisE7 with large apolar residues' results in loss of the water molecule and the resulting effect would be enhanced binding rates of both CO and 02 [43]. The Fe-CO complex is relatively apolar compared to the Fe-02 complex. As a consequence, replacement of HisE7 with aliphatic residues should increase Kco by eliminating the need for water displacement. In contrast, the Fe-Oz complex is highly polar and strongly stabilized by direct hydrogen bonding to HisE7 [13]. Apolar substitution at position E7 prevents this interaction and the net result is a notable decrease in 02 affinity. 5.2.1 CO ligated myoglobin and vibrational spectroscopy A large variety of CO ligated heme proteins and model compounds, studied by resonance Raman (RR) and infrared (IR) Spectroscopies, have shown an inverse correlation between v(C-O) and V(M-CO) [48-52]. The inverse correlation has been interpreted as resulting from back-bonding from the metal atom to the anti-bonding orbital of CO. Li and Spiro have suggested that a proton donor source near the oxygen atom of the bound ligand may enhance the degree of 132 A333 .3 .3a 5m .638 .52 .0 day: as .m .1820 ”.o a. .awa ”.< .m .235 8.5 “V8825: age §§e§§sa§8u§8§a§a8§§§o§s§ .338. u a. a. a. 8 $2. 3. $9.55 8.. a. Sod 8w 2. 84 :3. E .3935 8” ES 5 N m a ”85 8... $63.95 So a 585 8a m a. w E ~85 one Emrahm 8o : as... 82: 2. 8 v8... 3. Easing 8a. a. 9.3 SE. a a. 8m 83. 3. $9325 so a 38... 8o : 2. 8H 33. we .Saaaéfi 8a a :2 8a. m 8 o: :3 3w Aamsfim 8o 8 «3o 8.. m. ca 8» as... E 555535 25 a. «85 2: a. 8 2: a was 8 £93qu 8a a :3 8,. o a: 9: mad 3 $9.5m 83 m :3 8a. a. a 8" See a...“ 59538 :8 S :8... c8 3 o: 8" 83 we REEE o8 I :3 8° 2 .5 ca 25.: 3. E82?» 8... a ace 8». a S 9.. £3 a.“ 555.28 8» N «85 8” N 8 8 $2. 3 $925 82 $3 8... a 8 § 3.2 3. 8.555% 8: 85... 8e 3 a: 83 82. we EmEEm 8a. 8:. a: a a N8... 3 EmEeE 553:2 Ea: .m : $5 a 2 a as... 2.... 85a E 2 I S . a. mad 2.: ha a 2 3 S a 23 Se Ema as? .58» 25-33 .4 an}: p-23 {a A... 723 A723 is A... 723 532.. a: .3 an 8m 84 8a 133 back-bonding [50]. The increased back-bonding would also increase the carbon to metal bonding and the net result would resemble M5(+)=C=05<-> which would manifest itself in an increased v(M-CO) energy and a decreased v(C-O) energy. Although mutations of HisE7 resulted in large increases (Table l) in CO affinity, the IR spectra of all these mutants have v(C-O) bands in the 1960 to 1970 cm'1 region. The mutation of ValEll with threonine in porcine myoglobin decreased CO affinity 4-5 fold but shifted V(C-O) from 1945 to 1961 cm'1 [54] while the replacement of ValEll with Asn also decreased CO affinity but shifted v(C-O) from 1954 to 1916 cm'1 [50,53]. These results, as the previous results, suggest that there is little or no correlation between CO affinity, v(C-O) and Fe-C-O geometry. 5.3 Oxygen binding and hydrogen bonding As mentioned above 02 binds as an end on bent molecule, therefore, it is largely unaffected by steric hindrance. The iron-02 complex, unlike the iron-CO complex, is highly polar which makes it subject to stabilization by hydrogen bonding. Evidence for 02 being hydrogen bonded to HisE7 was reported in cobalt substituted proteins [55-58]. Evidence for the hydrogen bond comes from neutron diffraction and spectral studies including a deuterium isotope effect from vibrational spectroscopy. Studies with "hanging basket" model compounds, show that if the ether linkage is replaced by a proton donating amide linkage, the oxygen binding affinity increases by an order of magnitude while the CO binding affinity remains unaffected [27,59]. Confirmation of a hydrogen bond between the amide and the oxygen was provided by NMR studies [60,61]. Similar reports of enhanced 02 binding affinities have been reported for model hemes with distal alcohol or secondary amines [62] and phenyl urea substituents [63]. 134 Table 1 lists the CO and oxygen binding parameters for some myoglobin HisE7 mutants. The tabulated results show a strong effect of hydrogen bond donors in the distal pocket. When the strong hydrogen donating histidine ligand is replaced with any of the other listed ligands the oxygen affinity is decreased from one to three orders of magnitude. Many synthetic model compounds have been prepared that affect the CO and 02 binding affinities and subsequent spectroscopic properties. However, until recently there has not been a model that is ideally suited to studying the effects of hydrogen bonding to axial ligands. In chapter 3, using the stable oxovanadyl forms, we established the inu'amolecular hydrogen bonding of naphthyl Kemp acid porphyrin (NKAP). The usefulness of using stable molecules, as templates, for less stable system is well established [64-70]. In this chapter we will show the effects of hydrogen bonding on the vibrational modes of v(M-CO), where M = Fe(11) or Ru(II). 5.4 Experimental Raman spectra for the carbon monoxide bound complexes were obtained by using a 90° scattering geometry and were measured with a Spex 1877 monochromator equipped with a CCD detector. Laser emission at 406.7 nm was from a Coherent Innova 9O krypton ion laser. Optical absorption spectra, obtained by using a Perkin-Elmer Lamda 5 spectrometer, were recorded prior to and following Raman studies to confirm sample integrity. Anaerobic quartz cuvettes were used for both the Raman and optical absorption measurements for the carbon monoxide complexes. Raman spectra of the oxygen bound complexes were obtained by using a 180° scattering geometry on the detector described above. The sample was prepared by dissolving Co(III)KNAP or Co(III)NKAmP in methylene chloride 135 under an argon atmosphere. To the methylene chloride sample mixture was added an excess of water saturated with sodium dithionite. The dithionite concentration within the mixture was sufficiently high so that any free oxygen that might have been present was reduced to water. At the same time the Co(III)KNAP or Co(III)NKAmP compounds were reduced to Co(IDKNAP or Co(II)NKAmP, respectively. The aqueous and organic layers remained separated. The organic layer, consisting of methylene chloride and Co(II)NKAP or Co(II)NKAmP, was transferred via gas tight syringe into a specially prepared bulb that was attached, with heat shrink tubing, to a removable epr tube [71]. Prior to adding the sample dimethyl formarnide (DMF) was placed into the bulb and was carried through several freeze-pump-thaw cycles to remove any oxygen from the DMF. After placing the samme into the bulb we opened the system to vacuum which evaporated the methylene chloride but not the DMF. In this way we were able to degas and reduce the sample while maintaining an aqueous free DMF solvent. Once prepared in DMF the Co(II)NKAP or Co(II)NKAmP sample can be transferred into the epr tube that could be sealed with the heat shrink. A sample holder for the epr tube was previously designed by Tony Oertling [71,72] so that a stream of nitrogen gas could be blown though a dewer of liquid nitrogen and on to the sample. The flow rate of the nitrogen gas stream is used to regulate the temperature of the sample. In this way it was possible to reach temperatures as low as -120°C. 5.5 Results 5.5.1 Carbon monoxide bound RuNKAP Figure 5.1 a represents the high frequency RR spectrum of CO bound ruthenium naphthoic Kemp's esterporphyrin. Figure 5.1 b represents the high frequency RR spectra of CO bound ruthenium naphthoic Kemp's acidporphyrin. 136 Figure 5.1 c represents the high range RR spectra of CO bound ruthenium naphthyl Kemp's amideporphyrin. In all of these spectra we identify the usual porphyrin core vibrational modes. These modes are all insensitive to the structure of the Kemp's functional group. Figure 5.2 represents the corresponding low frequency RR spectra of these compounds. In the low frequency spectra we see a vibrational mode in the 560 cm'1 region that is sensitive to the Kemp's functional group. We identify this vibrational mode as the carbon-ruthenium vibration of the CO bound ruthenium Kemp's porphyrin complex. Comparison of the metal-carbon vibrational frequency shows that the acid form of the Kemp porphyrin has the lowest energy while the ester form has the highest energy. This suggests that the CO ligand is more linear in the ester case than in the acid. If steric hindrance is the major factor controlin g CO binding discrimination than we expect the opposite results since a methyl group is much bulkier than a proton. 5.3.2 Carbon monoxide bound FeNKAP Figure 5-3 represents the high frequency region RR spectra of CO bound iron(II)porphyrins. In Fig. 5-3a we can identify the usual vibrational modes that are associated with five-coordinated Fe(II) octaalkylpophylins [73]. Figures 5-3 b, c and (1 represent the the high frequency region RR spectra of CO bound iron(II) Kemp's ester, amide and acid porphyrins, respectively. The naphthyl substituted meso carbon lowers the D411 symmetry of the porphyrin with a notable effect on the totally symmetric V4 and V5 vibrations. The V5 mode is shifted to lower energy while V4 becomes split into multiple components. The V10 frequency of all four CO ligated Fe(II) porphyrins is located at 1635 cm'1 which is shifted down from the 1655 cm‘1 region for typical planar porphyrins. This down shift is consistent with previously reported domed or ruffled 137 Figure 5-1 The high frequency resonance Raman spectra obtained with hex = 406.7 nm of (a) carbon monoxide bound ruthenium naphthly Kemp's esterporphyrin (b) carbon monoxide bound ruthenium naphthly Kemp's acidporphyrin (c) carbon monoxide bound ruthenium naphthly Kemp's amideporphyrin. NSIT‘Y E RELATiVE INT 138 —— o) NKEPRUCO/CHZCIZ 406.7nm @3 mW b) NKAPRUCC/CHZCIZ 406.7nm @3 mW C) NKAmPRUCO/CHZCQ 406.7nm reng 1481 - 1139.3 58.1 \ 00" 4 o ' l I. who” [in MV\ ' M.,, l \, QAWMNAUW‘M J 7 '1 0,00 1 2100 1 4100 RAMAN SHIFT (CM—1‘) 1 58.3 .7 ‘ fifi-lsn Figure 5-2 139 The low frequency resonance Raman spectra obtained with lex = 406.7 nm of (a) carbon monoxide bound ruthenium naphthly Kemp's esterporphyrin (b) carbon monoxide bound ruthenium naphthly Kemp's acidporphyrin (c) carbon monoxide bound ruthenium naphthly Kemp's amideporphyrin. RELATIVE INTENSITY 140 0) NKEPRUCO/CHZCIZ 406.7nm @3 mW b) NKAPRUCO/CHZCIZ 406.7n. . @3 mW C) NKAmPRuCO/CHZCIZ 406.7nm @Smw I“ /\ ° (0 W J\/\ in C "NEVx‘V-AAJ M. ., Ib kw I kw MAW. _ 360 450 560 600 750 RAMAN SHIFT (CM— 1) 800 Figure 5-3 141 The high frequency resonance Raman spectra obtained with lex = 406.7 nm of (a) carbon monoxide bound iron(II) octaethylporphyrin in CH2C12 (b) carbon monoxide bound iron(II) naphthly Kemp's esterporphyrin (0) carbon monoxide bound iron(II) naphthly Kemp's amideporphyrin (d) carbon monoxide bound iron(II) naphthly Kemp's acidporphyrin. IR’ELATIVE INTENSITY ? O - 1024.9 -1 . g5 3 1 K1 ,.,-“#2:" 147. o) Fe(II)OEF’CO,/CH2C12 406.7nm @3 row b) OCFe(il)I\IKEP/’CHZC|2 406.7nm @3 mW c) COFe(II)NKAlrlP/CHZCIZ 406.7nm @3 mW d) OCFe(II)NKAP/CHZCI2 406.7nm @3 mW 1 2'00 1 Too RAMAN SHIFT (CM—- 1) d- to r\ I”) I 0? 31 00- l0 8 "I“: 001“) '2 10/1‘0 1'“?- I E: 1 600 {Fax .. Figure 5-4 143' The low frequency resonance Raman spectra obtained with ch = 406.7 nm of (a) carbon monoxide bound iron(II) octaethylporphyrin in CH2C12 (b) carbon monoxide bound iron(II) naphthly Kemp's esterporphyrin (c) carbon monoxide bound iron(II) naphthly Kemp's amideporphyrin (d) carbon monoxide bound iron(II) naphthly Kemp's acidporphyrin. RELATIVE INTENSITY O) FeUQOEPCO/CHZCIZ 406.7nm @3 .nw b) OCFe(II)NKEP/CH2C|2 406.7-om @3 mW c) OCEe(|I)NKAmP/CH2C!2 406.7nm @3 mW d) OCFe(II)NKAP/CHZC|2 406.?nm @113. mW “p fl..._=h 703 it b NW waI/wmxwwr’ II [AI W.,/“W _ \ \ . 5% I\ \ WEE/"M “\M’m/M‘N'mxh, “If” .31/ V/W‘u “MM/fix [\1 A E Show“ J \J \1 360 450 550 650 760 RAMAN SHIFT (CM—‘1) Figure 5-5 145 The low frequency resonance Raman spectra obtained with lex = 406.7 nm of (a) carbon monoxide bound iron(II)NKAP in CH2C12. Iron was inserted into free base NKAP in the form of iron(II)carbonyl carbonate. 'Ihe carbonate anion left the COFe(II)NKAP complex in the deprotonated five-coordinate form. (b) carbon monoxide bound iron(II)NKAP in CH2C12 and 2mL of HOAc that would leave the COFe(II)NKAP complex in the protonated five-coordinate form. NSIW TE RELATIVE IN 146 o) (CO)FeNKAP/CHZCIZ 406.7nm_@.3 mW b) (CO)FeNKAP/CHZCIZ/HOAC 406.7nm @3mW 284 ** 3 [\ pf) 1’) m RMJW r J WM\/f/k"fl { —u~¥ :3- :- 664 <1: 360 460 560 600 RAMAN SHIFT (CM— 1) 700 147 metalloporphyrins [74]. The domed or out-of-plane metal structure for the five- coordinated complex is consistant with the crystal structure of the oxovanadyl naphthyl Kemp's amideporphyrin [75] . Figure 5-4 represents the low frequency region RR spectra of CO bound iron(II)porphyrins. It was previously established that v(Fe-CO) for CO bound Fe(II)porphyrins is the range of 525 to 540 cm'1 [76]. In Fig. 5-4a we can assign the mode at 538 cm'1 to v(Fe-CO) for the unperturbed CO ligated iron porphyrin. In Fig. 5-4b we detect little, if any, effect on the the frequency of v(Fe-CO) for the Kemp's ester complex. If steric hindrance were a major influence on the frequency of v(Fe-CO) then we would expect a much greater influence from the ester complex. In Figures 5-4 0 and d we see that the v(Fe-CO) frequency is shifted by 7 cm'1 compared to the unrestricted COFe(II)OEP complex. Even if we allow one wavenumber for detection limits of the intrumentation we we have an obvious effect on the frequency of v(Fe-CO) for both the amide and acid complexes. Both of which are strong hydrogen bond donors [76]. Figure 5-5 a and b represents the RR spectra of COFe(II)NKAP in the deprotonated and protonated forms respectively. In the deprotonated form of COFe(II)NKAP we detectect the frequency of v(Fe-CO) at 537 cm'1 which is within experimental agreemant with the frequency of v(Fe-CO) for COFe(II)OEP. In Figure 5-5b we have added 2 [IL of acetic acid (HOAc) to the sample that was used to obtain Figure 5-5a. The shift to lower energy for the hydrogen bonding donor of the protonated COFe(II)NKAP complex is in agreement with the frequencies v(Fe-CO) reported in Figure 5-4d. 5.6 Discussion Table II is a summary of the observed frequencies of v(M-CO) M = Fe(II) or Ru(II) for "unhindered" (CO)M(II)OEP or the naphtyl Kemp's porphyrins. The 148 Kemp's acid and amide compounds are likely to hydrogen bond with CO whereas the Kemp's ester complex would likely act as a hindered structure. If steric hindrance is indeed the factor that most influences the frequency of v(M-CO) then we would expect the frequency for the ester to be the furthest away from the unhindered frequency of (CO)MOEP. However, this is not the case in the observed frequencies. In fact the greatest effect on the frequency of v(M-CO) is detectedinthe7cm-l shiftinbothNKAPanNKAmP. This is most likely explained by the hydrogen bonding effect. It may be true that the CO is tilted off axis from the linear M-C-O case but this can also be explained by an attraction of the CO by the adjacent hydrogen bond donors. This situation would then produce a two-fold effect on the vibrational frequencies. One effect would, indeed, be the tilting of CO and the other effect would be polarity induced by the hydrogen bond. Table II. The observed v(M-CO) {M = Fe(II) or Ru(II)} frequencies for OEP, NKAP, NKEP or NKAmP in methylene chloride. Raman excitations were obtained from lex = 406.7 nm. Compound Solvent v(M-OO) (CO)RuNKEP. CH2C12 564 (CO)RuNKAP CH2C12 560 (CO)RuNKAmP CH2C12 561 (CO)F60EP CH2C12 538 (CO)FeNKEP CH2C12 537 (CO)FeNKAmP CH2C12 531 (CO)FeNKAP CH2C12 531 (CO)FeNKAP CH2 C12 deprotonated 537 (CO)FeNKAP CH2C12 w/HOAc 533 149 The fact that NKEP has virtually no effect on the frequency of v(M-CO) suggests that steric hinderence is a minor factor affecting bound CO in these compounds. These results would also suggest that steric hinderence from the distatal histidine of hemeproteins is not an essential element in the proteins ability for ligand discrimination. The observed v(M-CO) frequency for the deprotonated form of NKAP would further suggest that polar interactions in themselves are also not a major factor affecting the ligand binding abilities of hemeproteins. Indeed, polar effects may be placed on the ligated form of the heme, however, they are most likely the direct result of hydrogen bonding from the distal histidine. 150 REFERENCES [l] Branden, C. and Tooze, J. Introduction to Protein Structure, Garland Publishing, Inc., New York and London, 1991 [2] Perutz, M. F. Aannu. Rev. Biochem, 1979, 48, 327-386 [3] Perutz, M. F. Trends. Biochem. Sci, 1989, 14, 42-44 [4] Perutz, M. F.; Fermi, G.; Luisi, B. and Liddington, R. C. Acc. Chem. Res. 1987, 209 309-321 [5] Perutz, M. F. Annu. Rev. Physiol. 1990, 52, 1-25 [6] Gibson, Q. H. Biochem. J., 1959, 71, 193 [7] Antonini, E. and Brunori, M. Hemoglobin and Myoglobin and their Reactions with Ligands, Elsevier Science Publishers B.V.; Amsterdam, 1971 [8] Collman, J. P.; Brauman, J. 1.; Iveson, B. L.; Sessler, J. L.; Morris, R. M. and Gibson, Q. H. J. Am. Chem. Soc., 1983, 105, 3052-3064 [9] Lavalett, D.; Tetreau, C.; Mispelter, J.; Momenteau, M. and Lhoste, J .-M. Eur. J. Biochem, 1984, 145, 555-565 [10] Traylor, T. G.;'1‘suchiya, S.; Campbell, D.; Mitchell, M.; Stynes, D. and Koga, N. J. Am Chem. Soc., 1985, 107, 604-614 [1 1] Traylor, T. G.; Koga, N. and Deardurff, L. A. J. Am Chem. Soc., 1985, 107, 6504-6510 [12] Traylor, T. G.; Koga, N.; Deardurff, L. A.; Swepston, P. N. and Ibers, J. A. J. Am. Chem. Soc., 1984, 106, 5132-5143 [13] Springer, B. A.; Egeberg, K. D. Sligar, S. G.; Rohlfs, R. J.; Mathews, A. J. and Olson, J. S. J. Biol. Chem, 1989, 264, 3057-3060 151 [14] Collman, J. P.; Brauman, J. J.; Halbert, T. R. and Suslick, K. S. Proc. Natl. Acad. Sci., 1976, 73, 3333-3337 [15] Marks, G. 8.; Brien, .1. F.; Nakatsu, K. and McLaughlin, B. E. Trends Pharmacol. Sci., 1991, 12, 185-188 [16] Vandegriff, K. D. and Olson, J. S. J. Biol. Chem, 1984, 259, 12619-12627 [17] Lemon, D. D.; Nair, P. K.; Boland, E. J.; Olson, J. S. and Hellmuns, J. D. J. Appl. Physio. 1987, 62, 798-806 [18] Perutz, M. F. Nature, 1970, 228, 726-734 £196 Norvell, J. C.; Nunes, A. C. and shoenbom, B. P. Science, 1975, 190, 568- 7 [20] Collman, J. P.; Brauman, J. J.; Halbert, T. R. and Suslick, K. S. Proc. Natl. Acad. Sci., 1976, 73, 3333-3337 [21] Hayashi, Y.; Yamada, H. and Yamazaki, 1. Biochem. Biophys. Acta., 1976, 427, 608-616 [22] Satterlee, J. D.; Teintze, M. and richards, J. H. Biochemistry, 1978, 17, 1456- 1462 [23] Choc. M. G. and Caughey, W. S. J. Biol. Chem, 1981, 256, 1841-1838 [24] Phillips, S. E. V. and Schoenbom, B. P. Nature, 1981, 292, 81-81 [25] Barlow, C. H.; Ohlsson, P.-I. and Paul, K.-G. Biochemistry, 1976, 15, 2225- 2229 [26] Shimada, H. and Caughey, W. S. J. Biol. Chem, 1982, 257, 11893-11900 [27] Momenteau, M. and Lavalette, D. J. Chem. Soc., Chem. Commun., 1982, 341-343 [28] Satterlee, J. D. Inorg. Chim. Acta.,l983, 79, 170-172 [29] Smith, M. L. Ohlsson, P.-I. and Paul, K.-G. FEBS lett., 1983, 163, 303-305 [30] Shanna, V. S.; Malayakel, J. E. and Waterman, M. R. J. Biol. Chem, 1982, 257, 11887-11892 [31] Stryer, L. Biochemistry, Freeman: New York, 1988, pp 148-150 [32] Peng, S.-M. and Ibers, J. A. J. Am. Chem. Soc., 1976, 98, 8032-8036 152 [33] Gerothanassis, I. P.; Momenteau, M.; Hawkes, G. E. and Barrie, P. J. J. Am. Chem. Soc., 1993, 115, 9796-9797 [354%] llgliriyan, J.; Wilz, S.; Karplu, M. and Petsko, G. A. J. Mol. Biol., 1986, 192, [35] Huber, R.; Epp. and Forrnanek, H. J. Mol. Biol., 1970, 52, 349-354 [36] Padlan, E. and Love, W. J. Biol. Chem, 1974, 249, 4067-4078 [367; Norvell, J. C.; Nunes, A. C. and schoenbom, B, P. Science, 1975, 190, 568- 5 . [328] Heidner, E. J .; Ladner, R. C. and Perutz, M. F. J. Mol. Biol., 1976, 104, 707- 7 2 [39] Quillin, M. L.; Arduini, R. M.; Olson, I. s. and Phillips, G. N., Jr. J. Mol. Biol., 1993, 234, 140155 [40] Kim, K. and Ibers, J. A. J. Am. Chem. Soc., 1991, 113, 6077-6081 [41] Kim, K.; Fettinger, J.; Sessler, J. L.; Cyr, M.; Hugdahl, J.; Collman, J. P. and Ibers, J. A. J. Am. Chem. Soc., 1989, 111, 403-405 [42] Quillin, M. L.; Brantley, J. R. E.; Johnson, K. A.; Olson, J. S. and Phillips, G. N., Jr. FASEB, 1992, 6, 2687 [43] Quillin, M. L.; Brantley, J. R. E.; Johnson, K. A.; Olson, J. S. and Phillips, G. N., Jr. J. Mol. Biol., 1993, 234, 140-155 [44] Brantley, J. R. E., Jr.; Smerdon, S. J .; Wilkinson, A. J.; Singleton, E. W. and Olson, J. S. J. Mol. Biol., 1993, 268, 6995-7010 [45] Hargrove, M. S.; Singleton, E. W.; Quillin, M. L.; Ortiz, L. A.; Phillips, G. N., Jr.; Olson, J. S. and Mathews, A. J. J. Mol. Biol., 1994, 269, 4207-4214 [46] Egeberg, K. D.; Springer, B. A.; Sligar, S. G.; Carver, T. E.; Rohlfs, R. J. and Olson, J. S. J. Mol. Biol., 1990, 265, 11788-11795 [47] Rohlfs, R. J .; Mathews, A. J .; Carver, T. E.; Olson, J. S.; Springer, B. A.; Egeberg, K. D. and Sligar, S. G. Proc. Natl. Acad. Sci., 1990, 265, 3168-3176 [48] Uno, T.; Nishimura, Y.; Tsuboi, M.; Makino, R.; Iizuka, T. and Ishirnura, T. J. Biol. Chem, 1987, 262, 4549-4556 [49] Yu, N.-T and Kerr, E. A. in Biological applications of Raman Spectroscopy, Spiro, T. G. Ed., John Wiley & Sons, 1988 Chapter 2 [50] Li, X.-Y. and Spiro, T. G. J. Am. Chem. Soc., 1988, 110, 6024-6033 153 EslgsNagai, M.; Yoneyarna, Y. and Kitagawa, T. Biochemistry, 1991, 30, 6495- [52] Li, T.; Quillin, M. L.; Phillips, G. N., Jr. and Olson, J. S.; Biochemistry, 1994, 33, 1433-1446 [53] Balasubramanian, S.; Larnbright, D. and Boxer, S. Proc. Natl. Acad. Sci. USA, 1993, 90, 4718-4722 [54] Smerdon, S. J .; Dodson, G. G.; Wilkinson, A. J .; Gibson, Q. H.; Blackrnore, R. S.; Carver, T. E. andOlson, J. S. Biochemistry, 1991, 30, 6252-6260 5515.7]4Yonetani, T.; Yamarnoto, H. and Iizuka, T. J. Biol. Chem, 1974, 249, 2168- [56] Phillips, S. E. and Schoenbom, B. P. Nature, 1981, 292, 81-81 [57] Hanson, J. C. and Schoenbom, B. P. J. Biol. Chem, 1981, 153, 17-146 [58] Kitagawa, T.; Ondrias, M. R.; Rousseau, D. L.; Ikeda-Saito, M. and Yonetani, T. Nature, 1982, 298, 869-871 [59] Lavalette, D.; Tetreau, C.; Mispelter, J.; Momenteau, M and Lhoste, J.-M. Eur. J. Biochem, 1984, 106, 5132-5143 [60] Mispelter, J.; Momenteau, M.; Lavalette, D. and Lhoste, J .-M. J. Am. Chem. Soc., 1983, 105, 5165-5166 [61] Gerothanassis, I. P.; Momenteau, M and Loock, B. J. Am. Chem. Soc., 1989, 11 1, 7006-7012 [62] Chang, C. K.; Ward, B.; Young, R. and Kondylis, M. P. J. Macrame. Sci. Chem, 1988, A25, 1307-1326 [63] Wuenschell, G. E. Tetreau, C.; Lavalette, D. and Reed, C. A. J. Am. Chem. Soc., 1992 114, 3346-3355 [64] Bajdor, K. and Nakamoto, K. J. Am. Chem. Soc, 1984, 106, 3045-3046 [65] Proniewicz, Z.L.; Bajdor, K. and Nakamoto, K. J. Phys. Chem, 1986, 90, 1760-1766 [66] Reczek, C.M.; Sitter, AJ. and Temer, J. J. Mol. Spec, l985?6?, [67] Schappacher, M.; Chottard, G. and Weiss, R. J. Chem. Soc., Chem. Commun., 1986, 93-94 154 [68] Gold, A.; Jayaraj, K.; Doppelt, P.; Weiss, K; Chottard, G.; Bill, E; Ding, X and Trautwein,A.X. J. Am. Chem. Soc, 1988, 110, 5756-5761 [69] Czemuszewicz, R.S. and Macor, K.A. Raman spectrosc, 1988, 19, 553-557 [70] Oertling, W.A.; Kean, R.T.; Wever,R. and Babcock, G.T. Inorg. Chem, 1989, 29, 2633-2645 [71] Kean, R Ph.D. Dissertation Michigan State University, 1987 [72] Personal communication with Tony Oertling where he described the ”back scatter" set up that was altered to accommodate low temperature spectroscopy. [73] Spiro,T.G. (Ed); Biological Applications of Raman spectroscopy, Vol.3, 1988 [74] Prendergast, K. and Spiro, T. G. J. Am. Chem. Soc. 1992, 114, 3793-3801 [75] The crystal structure for oxovanadyl and (CO)Ruthenium naphthyl Kemp's amideporphyrins have been determined. The manuscript is in preparation. [76] (a) Pauling, L. The Nature of the Chemical Bond, Cornel University Press, Ithaca, 1960 (b) Pimentel, G.C. and McClellen, AL. The Hydrogen Bond, W.H. Freeman and Co., New York, 1960