INFRARED 0-H ABSORPTION INTENSITY
Dec. 20, 105s
GAS9
TABLE IV THE REACTION OF n-C3F, RADICALS WITH HYDROGEN-DEUTERICM MIXTURES Run no.
110 112 128 129 130 131 132 133 131 135
Time, min.
264 180 121 120 120 120 120 120 90
90
7 -
Dz
Pressures, cm.Hz Ketone
n
4.74 4.17 4.12 3.96 4.10 4.03 4.14 4.12 4.13
3.02 1.07
0.98 0.99
1.21
TEmp., K.
6 4.55 4.21 4.02 4.11 4.12 4.12 4.13 4.17 4.07
513 502 485 485 526 526 542
=
388 373 1019 887 779 80 1 1548 1010 612 280
ks
0.391
0.249
0.811
,216
1.072
,257
1,158
.277
0.966
,290
0.35
This gives a value of 1.7 for E,' - E3 which checks favorably with the value of 1.5 kcal. (13.8 - 12.3) from the separate determinations. This agreement rules out the importance of the H Dz and D H1 reactions. The ratio of the frequency factors A3'IAa
800 100 936 35 306
/?a' -_
d(CsFIU)/dl d(CaFiH)/dt
the relative collision numbers for the two reactions, 2'12.
7 X 103/RT
+
2G 980 68 1152 52
542 575 570
Although only five determinations were made, within experimental error the results confirm the values already obtained. Using the method of least squares the value for the ratio of the rate constants is k 3 ' / k 3 = 1.32e-1
Mass spectrum peak heights 51 52
. 0.30
+ 5
A? 0.25
1.32 0.20
agrees with the independent value
I t is interesting to note that this value checks favorablv with that found bv Pritchard. et al.. for the CF3 hydrogen and ieuterium system, ' L e . , 1.45.8 ~~~h of these values are close to the ratio of
>ius
(8) G. 0. Pritchard, H. 0 Pritchard. H . I Schiff and A F. TrotmanDickenson, Tvans. Faraday SOL.,62, 849 (1956).
[COXCRIBUTIOX FROM
THE
1.7
1.8
1.9 I n ?1 7 . LU"11
Fig. 2.-Arrhenius
2.0
2.1
.
plot for the reaction of Hyplus Dy mixtures.
CaF,radicals
with
OTTAWA, OXTARIO,CAXADA
NOYESCHEMICAL LABORATORY, UNIVERSITY OF ILLINOIS]
Quantitative Correlation of the Infrared 0-HAbsorption Intensity in Aliphatic Alcohols with Substituent Constants BY THEODORE L. BROWN RECEIVED JULY 3 , 1958 A linear relationship between the infrared 0-H intensity and polar substituent constants, u*, has been found. T h e relationship is obeyed with good precision for fourteen aliphatic alcohols of varying structure. Deviations from the linear relationship are found for compounds in u.hich internal hydrogen bonding is probable. The application of the relationship to determination of new values of c*,and as an aid in determination of structure is discussed.
Introduction Studies of the infrared absorption band due to 0-H in aliphatic alcohols have led to the conclusion that the intensity of the band is highly sensitive to the nature of the groups attached to the carbinol carbon.'S2 I t has been found that there is a quantitative correlation of the intensity with the sum of the polar substituent constant^,^ g*, of the (1) T. L. Brown and M. T. Rogers. THISJ O U R N A L , 79, 677 (1957) (2) T. L. Brown, J. M. Sandri and H. H a r t , J . Phys. C h e m . , 61, 098 (1967). (3) R. W. T a f t in "Steric Effects in Organic Chemistry," M. S.
groups attached to the ~ a r b i n o l . ~In the present paper the results of further nieasurements on alcohols are reported. The purpose of this study is (a) to improve the precision of the correlation, (b) to correct or verify the results for methanol and 3chloro-1-propanol, which were open to question, and (c) to uncover the cause or causes of deviation from the common behavior. The variation in intensity throughout a series of Newman, Editor, John Wiley and Sons, Inc., N e w York, X. Y., 1956. Chapter 12. (4) T. L. Brown, Chem. Reus., 68, 581 (1958).
1’01. so
THEUDORE L. BROWN
8490
compounds possessing a common functional group has been related in several instances to the substituent constants characteristic of attached groups. Functional groups studied include cyanide,”6 carbonyl,’ aromatic OH8,9and aromatic NHz.lotll
Experimental Tlie procedure employed is similar to that previously deacribed.l -% Perkin-Elmer Model 21 spectrometer fitted with sodium chloride optics was employed. The cell thickness was 1.00 mm., the slit spectral width was calculated as 15.5 cm.-l; the integration was carried out numerically over an interval of 150 cm.-l. -1concentration range of 0.080.01 molar was employed for the solutions. .I11 of the alcohols employed were obtained coriimercially and were carefully purified prior t o use. Reagent grade carbon tetrachloride was dried with anhydrous magnesium sulfate prior to use as solvent. T h e value of intensity obtained for each of the alcohols studied is listed in Table I . Tlie values reported are those calculated from t h e measured band areas and do not contain :my correction for wing absorption .12 L’alues for methanol :tnd 3-chloro-1-propanol were reported previo felt that the value for methanol might be low, cause of possible sampling error caused by the compound’s volatility. I n the present work special precautions were taken t o avoid this source CJf error and the value obtained is 0.05 intensity unit higher than that previously obtained .I Because the intensity for 3-chlorcJ-l-propano1 seemed unusual1:- high, this compound also was restudied, using a different sample. The result, however, is precisely that obtained previously.’
closely similar (u* for ~iiethjd,hydrogen and phenyl are O.OU, 0.49 and 0.60, respectively). X value of 0.40 is therefore chosen for H,C=CH. HCEC: u* for the C&C-C group is 1.35. The \ for the H C r C group is undoubted1:- a little less than t h value of 1.30 has been chosen. (CH3)KSO.: u* for C H ~ K O ~ C H is Z0.50. With one less methylene group the value is increased t o about 2.8 times this value,3 or 1.30 for CHJiO?. (This compares well with 1.30 for CHzCN, a similar group.) Substitution of two methyls for hydrogen lowers this by about 0.2 (the difference in u* for ethyl and t-butyl is 0.20), so that u* for the (CH,),C K 0 2group is 1 .lo. -1lthough the values of u* estimated as above are nut :is precise as the established values, they are almost certainly within 0.1 of the correct value, and this order of precision is all that is required for the purpose a t hand. Using t h e estimated and established values of u*, the intensity is graphed in Fig. 1 u s . the sum of the a*-values for the three groups attached to carbinol, utilizing the intensity results for seventeen alcohols (ref. 1, 2 and Table I ) . The alcohols not identified on the figure are, in order of decreasing intensitj-: trichloroethanol, tribromoethanol, ethylenechlorohydrin, propargyl alcohol, 3-butyn-2-01, benzyl alcohol, 2-methyl-3-butyn-2-ol, methanol, allyl alcohol, 71propyl alcohol, sec-butyl alcohol, t-butyl alcohol, triethylcarbinol and tri-isopropylcarbinol.
Discussion of Results
Correlations of absorption intensity with molecular reactivity parameters are largely empirical, and there is no general basis as yet on which t o choose a suitable function for the variables. In TABLEI VALUESOF ISTENSITY FOR THE IXFRARED 0-H .~BSORPTIOX some cases log A has been correlated with substiti~ SOME-$LIPHATIC ALCOHOLS,MEASUREDI X CARBOX uent constants, in other instances simply the intensity itself seems t o afford meaningful results. I n TETRACHLORIDE the present study it does appear that the intensity Cornp~iund Intensity“ rather than its logarithm is best correlated with u*. Methanol 0.50 Inspection of Fig. 1 shows that for fourteen of 3-Chloro-1-propanol . 63 the alcohols the relationship between u* and in1,3-Dichloro-2-propanol . i i tensity is linear. The deviations of the other three 2- Nitro-Z-methyl-l-propanol .87 compounds are well outside the range of experi2-Chloroethanol . GO -mental error and will be discussed below. The lin2,2,2-Tribromoethanol 15 earity of the relationship is strikingly good; the Furfuryl alcohol .56 standard deviation from the line is less than 0.01. Tetrahydrofurfuryl alcohol . 60 The qualitative significance of the observed variIn units of 1 X 104mole-‘ 1. cm. -*. ation of intensity with changing electron-withCorrelation of Intensities with Values of u* .-The alcoliols drawal power of the attached groups has been disare considered as compnunds of the form R I R ~ R ~ C - O HThe . cussed previous1y.l 0-H intensity is to be compared with the suni of the u* In a recent study of the 0--H absorption in values for the three R groups. l-alues of u* for a number of aliphatic alcohols Fletti3has pointed out the prevafunctional grvups are given by Taft.3 I n a few instances where a value is iiot available it can be estimated by com- lence of internal hydrogen bonding, as evidenced parison with similar groups. This lias been done f i r the fol- by the frequencies of band maxima, and semi-quanIowinp. . . ~-~~ romnoiinds. -....=. ..~~... CBr8: The value of o* for CHzCl is 1.05, that for CH2Br titative estimates of intensity changes. The 0--H intensity is known to be extremely sensitive to hyis 1.00. The difference in u* for CH,F and CHeCl is 0.05, while for CHF2 and CHC12 the difference is 0.11. This is drogen bonding and generally increases markedly good indication that the differeuce in u* for the CCl, and when hydrogen bonds are formed.14,15 It seems c B r Bgroups should be roughly three times larger than- that quite probable that very weak hydrogen bonding for C H X I arid CHeBr, which is 0.05. -Accordingly, since should be detectable by accurate intensity measthe u* value for CCliIis 2.65, it is taken as 2..50 for CBr3. H2C-.=CH: u* for CHBCH==CHis 0.36, for CsH,CH=CH urements before any noticeable change occurs in it is 0.41; substitution of hydrogen for phenyl in the latter the frequency or gross band shape. group should result in 011ly a very slight change in u*, since The abnormally high intensity for 3-chloro- 1the elcctron-withdrawal properties of the two groups are propanol and 2-nitro-2-inethyl-l-propanolcan be ( 5 ) 11.I < . blander and H. \T. Thompson, Z ’ Y Q ? Faraday ~~. Soc.. 63, explained as due to intrainolecular hydrogen bondi.LO2 (1937). ing, with formation of a six-membered ring. (1;) T. I.. H r , , \ v n , ’ ~ I I JI D~ I ; R ~ ; A ~80, .. 7!1L (1!158). ( 7 ) €1 W.Thompson, I < . W. S r e r l h a m and 11. Jarnesim. . S ~ c < l ~ c ~Hydrogen bonding to other than first row elements possessing lone pairs is known to be relatively weak. (X) 1’. J . Stilne a n d 11. \V, ‘I’hornpx)n, ibid., 10, 17 (l!l.j7). Further, the formation of five-iiieinberetl rings b y (!I) ‘ I . I.. Brown, J . P h y s . Cheiiz., 61, 820 (1957). t13) & I , Si. c.Iilett, S p ‘ Y t i o c ! l l i i l . A C l ~ l ,10, “1 ( (10) S. Califano and R. XIoccia, GQZZ. chim. iLaZ,, 87, jS (1!1:7). 11 1) H. Tsuhornura. .r. Chcm. Piiys., 2 4 , 027 (19 (11) P. J. Krueger and H. W.Thompson, Proc. R o y . SOC.( L o i t d o ~ i ) . 2 4 3 8 . 113 119.ii). (1.5) C .11. HriKKini ‘and r.,C . I’irnentel, . I . P h y s . Chiin , 6 0 , I 1 i l . i -c
( 1 2 ) 1). .A. I
