ular weight as determined by light scattering. Although the identity of the combining species has not been determined, we will tentatively assume for simplicity in this discussion that it is the HgClz molecule. Katz further found that the addition of C1- or CN- would completely reverse the complexation leaving free DNA. He interpreted the large increase in the molecular weight that was brought about by the addition of HgC12, as being the result of partial aggregation. However, he assumed that the HgC12-DNA complex had the same refractive index increment (dnldc) as free DNA, while i t is probable that the complex has a higher dn/dc value than the free DNA. This would cause his reported molecular weights of the complex to be somewhat too high. n'ith this in mind, partial aggregation need not be postulated. Further investigation of this reversible complexing reaction has revealed that the ultraviolet absorption spectrum of DNA is substantially altered by the addition of HgC12. In 0.40 M acetate buffer, p H 5-6, free DNA has an absorption maximum a t 258 mp. On the addition of HgC12 (dissolved in the same buffer) the absorption maximum shifts over to 275 mp, The addition of NaCl 3000 3500 4000 4500 to this solution will cause the absorption maximum Wave length, A. to again shift back to 258 mp. This, together with Fig. 2.-Absorption spectra of absolute ethanol solutions Katz' observations, is strong indication that the of typical chelates: A, Sc( C9HeNO)s*C~HeNOH,1.2 mg. DNA-HgC12 complex is reversed by the addition Sc/l.; B, same as A with 10% water by volume; C, Sc- of NaC1. ( C Q H ~ C ~ ~ N1.2 O )mg. ~ , Sc/l.; D, same as C with 20% On examining the shift of the absorption specwater by volume; E, Sc( C~H~Cl~NO)~~CgH~C12NOH, 1.2 trum more carefully, we find that all curves go mg. Sc/l.; F, same as E with 20% water by volume. through an isosbestic point located a t 238.5 rnp (see Fig. 1). Defining r = total moles of added ClzN0)3: 0.60-1.20 mg. Sc/l., kav. a t 3430 8. = HgC12/moles P, we find that all curves for which 148.9, ha". a t 3970 8. = 191.3; Sc(CgH4C12NO)r r is less than 0.60 pass through isosbestic points a t CSHIC~ZNOH: 0.60-1.2s mg. Sc/l., k a v , a t 3430 A. 238.5 and 262.5 mp. On increasing the mercury = 201.1, kav. a t 3970 A. = 220.6; Sc(C9H4Br2)r concentration from r = 0.60 to 10, the absorption C9H4Br2NOHl3: 0.60-1.20 mg. Sc/l., kav,a t 3430 8. curves display a new isosbestic point a t 2T4.5 mp = 217.2, kav. a t 3990 A. = 220.2; ? C ( C S K ~ C ~ I N O )but ~ , still pass through the first one a t 238.5 rnp. 0.60-1.20 p g . SC/l., k a v . a t 3470 A. = 160.3, k a v . This behavior suggests in this particular case a t 4020 A. = 184.4; S ~ ( C ~ H ~ C I I N O ) ~ C Qthat H ~ Cthe ~ - reaction proceeds by a t least two steps INOH: 0.5Q-l.OOmg.S~./l.,k,,.at3470A. = 215.3, and that the first reaction is essentially complete before the second reaction begins. If this is the kav, a t 4020 A. = 210.2. Acknowledgment.-Funds received from an E. I. case, the curves in Fig. 1 which pass through the du Pont de Nemours and Company Grant-in-aid isosbestic points a t 238.5 and 262.5 rnp can be confor partial support of this investigation are grate- sidered to result from two components in equilibrium with each other and absorbing ultraviolet fully acknowledged. light independently. (13) Sc(CoH4BmNO)rwas too difficultly soluble in ethanol to permit Choosing two appropriate wave lengths, i t is study. possible to write two simultaneous equations for NOYESCHEMICAL LABORATORY the total optical density at these wave lengths in OF ILLINOIS UNIVERSITY URBANA, ILLINOIS terms of the concentration of the free DNA and the concentration of the complex. I n this case we chosen wave lengths of 257.5 and 271.0 The Interaction of HgClz with Sodium Thymo- mp have because they are near maxima, and yet as far nucleate from isosbestic points as possible; the results do BY C. A. THOMAS* not depend on this choice. RECEIVED MAY8,1954
D~~~ =
EE,; cDNA
+ ,ymplwx 2~51.6
Ccomplex
An interesting reaction of sodium thymonucleate (DNA) was investigated by Katz,l who found that HgCl2 undergoes a reversible combination with DNA which results in a large increase in the molec* Lilly Research Labs., Indianapolis, Ind.
D~~~ =
E E ~ A cDNA
+ p o271 mplex
Coomplex
( 1 ) S. Katz, THIS JOURNAL, 74, 2238 (1952).
Employing Chargaff's value of 6650 f. 50 (at 259 mp in the presence of salt) for the extinction coefficient with respect t o phosphorus (E(P)),2i t is (2) E Chargaff and R . Lipshitz, i b i d . , 75, 3658 (1953).
NOTES
Dec. 5 , 1954
6083
260 270 280 290 300 310 Wave length, m,u. Fig. 1.-The influence of HgC12 on the ultraviolet absorption spectra of DNA. D is defined as the “optical density” or log,, I,,/I. Spectra of DNA in 0.40 M acetate buffer, PH 5-6 with increasing r, r = moles HgC12/grundmoles DNA. 240
250
possible t o calculate all of the other extinction All of these observations are in accord with the coefficientsin the above equations. Solving for the decrease in size of the DNA molecule on the addiconcentration of the DNA-HgC& complex, we tion of HgClz observed by Katz. Discussion.-Since the purine and pyrimidine find that 0.56 f 10% moles of HgClz complex per mole of DNA phosphorus. Furthermore, the residues of DNA are primarily responsible for the equilibrium is well in favor of the complex form. absorption maxima a t 260 mp, the shift in the abThe stoichiometry of the second reaction cannot sorption spectra brought about by the addition be measured in this manner because the shift is of HgClz strongly suggests that the binding occurs too small. But i t is clear that the equilibrium a t these bases. Although the nature of this combination is unknown, i t is possible that the HgClz constant is not as large as in the first reaction. The character of the DNA precipitate which is coordinating with the conjugated double bond results from the addition of ethanol to DNA systems in guanine, cytosine and thymine much in solutions in 0.40 M acetate buffer, depends very the same manner that HgClz can combine with markedly on the amount of HgClz present. I n the mesityl oxidee3 Another possibility is that the absence of HgC12, the ethanolic precipitate is HgClz combines with NHz groups in adenine, guanfibrous and can be spooled up about a stirring rod. ine or cytosine. I n the Crick-Watson4 model for When 0.5 mole of HgClz have been added per each DNA, these bases are involved in hydrogen bonds mole of phosphorus, the character of the precipitate which hold the two polynucleotide strands tois a h e powder which settles slowly. X-Ray gether. It is likely that when these bases combine diffraction patterns of these two types of pre- with HgClz certain hydrogen bonds between the cipitates reveal that the HgC12-DNA precipitate polynucleotide chains will be destroyed. This is very amorphous while the ethanolic precipitate destruction would result in a more flexible molecule of the original DNA is highly crystalline, displaying which would then spontaneously assume a more 6 well-defined diffraction rings. compact configuration. This is in agreement with Preliminary experiments show that the sedi- the decrease in size observed by light-scattering‘ mentation constant (measured with about 0.2% and viscosity measurements. Since the original DNA in 0.40 M acetate buffer p H 5-6) increases in size and shape of the DNA molecule is restored a linear fashion with increasing amounts of HgC12. after the HgClz is removed, a sufficient number of For instance, when r = 0, sz0 = 13. On increasing hydrogen bonds must remain intact in the presr to r = 3, s20 becomes equal to 50. The intrinsic ence of HgClz to provide a “skeleton” to guide the viscosity of the complex a t an average gradient of reformation of the molecule when the reaction is 1000 sec.-l is about 9 compared to a value of 20 for reversed. It is highly probable that there is some interthe original DNA under the same conditions. The apparent gradient dependence of viscosity of (3) J. Chatt, Chcm. Reus., 48, 7 (1951); see pp. 11. the original DNA is decreased by a factor of 5. (4) J. D. Watson and F. H. C. Crick, Nature, 171, 747 (1953).
6084
NOTES
action between the phosphate groups and the mercuric ions Hg++ and HgC1+ about which we have no information. Acknowledgment.-The author is grateful for the help and advice of Dr. Sidney Katz. GIBBSLABORATORY DEPARTMENT O F CHEMISTRY HARVARD UNIVERSITY CAMBRIDGE, MASS.
The troublesome features of (1) and (2) are largely obviated in (3). Activated oxygen produced may react by 0 2 *
followed by
+ 02 + + 0 0 3
O+O~+M---tO~+M
The Photochemical Formation of Ozone. Foreign Gas Effects on the Mercury Sensitized React+ at 2537 and the Unsensitized Reaction a t 1849 A.
A.
BY DAVIDH.VOLMAN RECEIVED AUGUST16, 1954
+ +Hg + 20, AH = 5.5 kcal. + +HgO(g) + 0, AH = -87 + X Hg* + 0 2 +Hg + 0 2
(4) ( 5)
Equation 5 is common to mechanisms involved in (l), (2) and (3) since 0 atoms are postulated for all three. I t was felt that a study of the effects of foreign gases in the reaction system would give information relative to the mechanism of ozone production since foreign gases would play a role in the following processes Hg*
Following the formation of Hg(3P1)atoms by adsorption of mercury resonance radiation, 2537 A,, three processes ultimately leading to ozone formation may be considered‘,‘ Hg* Hg*
VOl. 76
+ M -+
Hg
+ M*
(6)
O?*+ M +0 2 + Ll* (7 ) as well as in eq. 5 above. In the course of the experiments, it became evident that information on the effects of foreign gases on the unsensitized reaction would also be desirable.
(1) (2)
Experimental The experimental flow method used was similar to that 02* (3) described earlier.’ High purity commercial gases from cylinders were measured by flow meters and dried over magnewhere X is the heat of sublimation of HgO(s). The sium perchlorate. The light source was a mercury-rare calculations are based on H g ( 3 P ~atom ) but the es- gas low pressure discharge tube in the shape of a helix, 70 sential features would not be changed by consider- mm. across and 250 mm. long. The light emitted from such ing instead H g ( 3 P ~ ) .The activated oxygen species a lamp is primarily a mixture of the 1849 and 2537 A. mermay be either normal oxygen molecules in high vi- cury resonance lines. Since the 1849 A. line will not be transmitted through a few mm. of liquid water: it is posbrational states or electronically excited oxygenS3 sible to prevent the short wave length radiation from reachCalculations from collision theory based on the ing the reaction zone. For the unsensitized reaction, the endothermicity of eq. 1 indicate that only one colli- lamp and reaction tube, 30 mm. diam., were both surrounded sion in about lo4would be effective. Since the quan- by air. Under these conditions ozone was formed in the abtum yield for the reaction is of the order of 0.03,4 sence of mercury vapor. For the sensitized reaction, the vessel, 20 mm. diam., and surrounding lamp were and since long chains are not possible, i t appears reaction contained in a water-bath thermostated a t 45O, the temunlikely that this reaction is an important one. perature for optimum light intensity.8 Under these conThe value of X for eq. 2 is not known but Noyes5 ditions ozone was not formed in the absence of mercury vapor but was formed in the presence of mercury vapor. The estimates that it is not over 23 kcal. Thus (2) pressure of mercury vapor entering the reaction zone was is exothermic, and therefore (2) and (3) are both adjusted to a value corresponding to saturation a t 20’ in the possible paths. Dickinson and Sherrilla have manner described previously.7 Since oxygen gas a t atmosshown that a t least 7 molecules of ozone are formed pheric pressure can absorb only negligible amounts of light for each mercury atom passing through the reaction of wave length 2537 A., the formation of ozone in the unreaction may be attributed primarily to the abzone. Volman4 has evidence that this value may sensitized sorption of the 1849 A. resonance line by oxygen molecule. actually be considerably higher. Since the forma- Ozone in the effluent gases was determined iodimetrically tion of ozone from oxygen is 34 kcal. endothermic, after absorbing in a neutral potassium iodide solution. 0 2
a single activated mercury molecule can a t best account for 3 ozone molecules. This could possibly be as high as 4 if the heat evolved in the formation of HgO were available. Even these yields are unlikely since mechanisms giving these values do not appear probable. A postulated solution to this dilemma is the reaction3 HgO(g)
+ 0 2 +Hg +
0 3
However this reaction is endothermic to the extent of 684 kcal. (1) Other processes involving postulated but not established molecular species have sometimes been considered, i.e., HgOt. Cf. W. A. Noyes, Jr., THISJOURNAL, 49, 3100 (1927). (2) Thermochemical values from F. R. Bichowgky and F. D. Rossini. “The Thermochemistry of Chemical Substances,” Reinhold Publ. Corp., New York, N . Y.,1936. (3) W. A. Noyes, Jr., and P . A . Leighton, “The Photochemistry of Gases,” Reinhold Publ. Corp., New York, N. Y.,1941, p. 225. (4) D. H. Volman, J . Chcm. Phys., 21, 2086 (1953). JOURNAL, 13, 514 (1931) (5) W. A Noyes, Jr., THIS (6) R . G. Dickinson and M. S. Sherrill, Proc. Natl. Acad. Sci., l a , 175 (1926).
Results and Discussion Number of Cycles Involving Mercury Vapor.I n the sensitized reaction a red-orange deposit of mercuric oxide was observed in agreement with the results of Dickinson and SherrilLB This deposit was heaviest a t the entrance to the reaction zone and tapered off so that no deposit was observable after about 8 cm. The effluent gas was found to be free of mercury vapor. From the yield of ozone and the vapor pressure of mercury entering the reaction zone, a value for the number of ozone molecules formed for each mercury atom entering the reaction zone may be determined. For example, a t a flow rate of 1.O liters oxygen per minute, the yield of ozone was 4.2 X loF6 mole per minute. The gas stream contained mercury equivalent to saturation vapor pressure a t 20.0’. Thus for each mercury atom about 60 molecules of ozone were (7) D. H. Volman, J . Chcm. Phys., 1 4 , 707 (1946). (8) L. J. Heidt and H. B. Bayles, THISJOURNAL, 73, 5728 (1951).
