refraction of the nicnolayer is assumed to be the same as that for solid acid, the observed reflectance changes agree with the calculations, provided the film thickness is taken to be about 3 A. less than the molecular chain length of the solid. Reduced film thickness may be due to immersion into the water of the hydrophilic group a t the end of each chain, or to tilt each chain away from an axis perpendicular to the water surface. The reflectivity change caused by a stearic or palmitic acid monolayer is sensitive to the purity of the water. This probably is due to a change in film
thickness and possibly to a change in refractive index. Interference theory accounts satisfactorily for the reflectivity change increase with decreasing wave length and for the reflectivity change a t high angles of incidence. Acknowledgments.-We are greatly indebted to Dr. Katherine Blodgett for her helpful advice on the preparation and properties of fatty acid monolayers, and for critical reading of this report,. We also wish to thank C. P. Goody for his assistance in construction and operation of the equipment and in calculation of the results.
STUDY OF DEHYDRA4TIOSOF MONTMORILLOSITE AND VERMICULITE BY ISFRBRED SPECTROSCOPY BY J. J. FRIPIA4T,*J.
CH.4USSIDOX** AXD
R. TOUILLAUX*
*Labmatoire des colloides des sols tropicaux, I.N.E.A.C., lnsfitut Agron~omiqziede I'UniversitB de Loicvnin (Belgique), ctnd **Centre Nntional de la Recherche Agronomique, Versailles (France) Received JanunrU 1 8 , 1960
Dehydration processes for homoionic montmorillonites and vermiculites have been investigat'ed by studying variations of dOOl spacings (by X-ray diffraction), intensit,y of angular vibration band of water ( a t 6.1 p ) and inteneity and composition of OH vibration band (2.7 to 3.5 p ) , in relation to dehydration temperature. Experimental results show t h a t : 1. Wat,er molecules remain u p to a final stage of dehydration, and even after the collapse of the interlayer space. 2. Lattice dehydroxylation start's before dehydration is achieved. It has been proved that, as dehydration progresses, the stretching vibration frequency of water hydroxyls unexpectedly increases progressively, since the collapse of layers should decrease OH. . . O distance and strengthen hydrogen bond. A theoretical interpretation is proposed, which consists in introducing in Schrodinger's equation, an additional potential energy originating from the application of fields existing inside the interlayer space, to the OH dipoles. I t is assumed that molecules remain in this internal space after the apparent collapse of layers and indirect proofs of this are given.
The dehydration of inontniorillonite and verMoreover, t,he structures as they are now known miculite has been investigated by means of iiifra- show the absence of surface hydroxyl groups, as red spectroscopy using two absorption bands : far as the interlayer space is concerned. This the first one corresponds to the angular vibration fact greatly simplifies the interpretation. of the water molecule a t 6 p , the second to the 0-H Therefore the purpose of this study is to investistretching vibration of H 2 0 between 2.7 and 3 . 5 ~ . gate the relations between characteristic features This band covers a wide frequency range since of the OH vibration band of hydration water molevarious hydrogen bonds may be formed between cules and the basal spacing. Iu order to point the hydroxyl group and neighboring oxygen atoms. out the influence of cations, several homoionic Lattice hydroxyls are characterized by vibration clays have been studied. frequencies depending upon the crystal structure. I. Experimental Methods For vermiculite, one vibration band only occurs A. Samples.-The fraction smaller than 2 p of the Camp at 3704 em. -I, whereas for montmorillonite two bands may be observed, at 3745 and 3650 em.-', Ikrteau Montmorillonite and of a Transvaal Vermiculite was separated by centrifuging and s:tt'uratcd by chlorides oc respectively. Yr,Li, Na and K. The study by infrared spectroscopy of the dcThe main propertics of these cluj-s :w suniin:trizt:d in hydration of silica gels and glass powders has al- Table I. ready been the subject of many papers. Kiselev,' TARLE I Renesi and Jones,z Young3 haye shown that as the coverage by water and the surface density of C A T I U S EXCH.4h'GE C A P A V I l ' Y (B.k;.c'.), S P E C I F I C YLRFACl.; (80)A X D ClI.4lZGE 1)E.USITY (g) ( I F TIIE CLAYS HTI.DIE1) silanol groups decrease, the maximum of the abs sorption band shifts toward higher frequencies. (i*lcctroiis/ IIliL2j This fact is easily interpreted by the increase of -,>the average O H . . . . 0 distance. Montmorillonite 05 1-2 0.8 It is interesting t o study dehydration processes Vermiculite 130 ti00 1 3 for expanding lattice clay minerals, where the space The base exchange capacity is detc~minedby drsorbiny occupied by water molecules can be defined by ammonium from clays saturated with this cation a t pH 7 . basal spacings (d001) measurements. The specific surface, SO, is measured by ethylene glycol (1) A. 1'. Kiselev and V. I. Lygin, Second Intern. Congress of Surface 11. Solid-gas interface, ButterTFortlis, London, 1957, p .
Actir i t g 204
(2) H A Bents1 and A . C Jones. Trris ,JoL.R% 4 ( 3 ) G . 3. Young J Colloid Sci., 13, ti7 (19.58)
~
63, , 179 (1959).
retention according to a variant of t,he Heridricks and Dyal mrthod . 4 The chemical composition^ uf thr rla!-s arc ______._
(4) S. B. llendricks and L. A . Dyal, Soil. Sci..69, 421 i l Y S 0 ) .
1235
Sept., 1960 ,
1.2
1
100
2800 5.5 5.i5 6.0 6.25 6.5 p \\-ave lengt,hs:,cin. -l. Fig. l.--T>yical spectra obtained for Li-montmorillonite: A, ahsor1)aiice of the OH stretching vibration (lattice iwat,er) band, as recorded z'ers1L.s the wxve number for several dehydrat>iontemperatures; H, absorbance of the OH (water) stretching vibration ziersIis the xave number for several dehydrat,ion C, angular vibration band of water. Transmission ( c;i:)uersus wave length ( p ) for several dehydration temperatures. Iempcratures: -, 20" : 05"; -'-, 175"; . , , ,, 250': , 300". 3(300
1000
3200
-)
Tlie location of the 001 reflection is noted arid a correctioii made in order t o take account of the slightly lowered position of the clay film in t'he diffractometer. Vermiculite: From infrared spectra, transmittances are converted into absorbances and the curve obtained between 2.5 and 4p Si6,444 + ( A13+,Fe3'jl.j6] I V [ ( a 1 3 + I F e 3 + ) ~ . 2 ~ ( ~ I g * + , Fv10ao(OH)4 e 2 + ) ~ . 7 ~ ] in this new scale is drawn. For vermiculite, it can be seen that the curve is dissymmetric in respect to the wave length The samples hthve been used as films, tho s w f a r e wight, axis; the absorbance a t 2.5, is always stronger than thc of which was about 80 mg./cm.2. These films itre obtained absorbance observed at, 4.0~;this probably is due t o difby slow evaporation of the suspensions under Yaciium. fusion initiated by the distribution of particle sizes.5 B. Apparatus.-The film is placed on a 2 X 1 m i . plttti- A straight line is drawn betxeen points representing abnum screen arid pressed between two copper plates. The sorbances a t the limits of the band, and measurements of bottom plate holds a heating wire and a thernioc~ouple. absorbance a t intermediate wive lengths are taken as tho It is thus possible t o heat screen and film lip t o 400", measure differences between the curve and the straight line. For thc temperature and keep it almost constant. montmorillonite the curve is distinctly more symmetric; The sample holder niay be introduced in oiie of Ilic lieitins tlie same method of calculation. however, is used. of the I R 4 Beckman spectrograph, fitted with ritlciuni Figure 1 gives a example of calculation for typical spectra. fluoride optics, while an enipty screen is introtiuced in the I n the case of the angiilar vihration band a t Cip, it is imreference beam. possjble to operate in ,a similar way. Clay minerals conThe same holticr (wi lie used in t,hc goniometer of a 1 1 ining m;tgnesiuni give :t hand of v:triablr, intensity, X-ray Philips diffractoinctcr (Cu tube). both bands ovrrlal) ntered approximately at 6.8,; Settings of both instr ents were as follo\vs: 1124: partially. For this reason the measurement of absorbanrc s eed, 0.25 ,/Inin., gain 3 period 2", slit, 2 X standard for the mater band was taken as the difference between Philips diffractomefe slit 1'; rate meter X 4 or X 8; absorbance a t 5.6p and absorbance a t the minimum of period 16" or 8"; multiplier X 1; speed'/r"/min. transmittance curve, t h a t is to say a t about 6 . 1 ~ . C. Procedure.-Two similar films are used in each exThe difficulty in getting "good" films of vermiculite must periment; one for infrared and the other for X-ray study. be emphasized particularly as compared with the ease of The first recording is obtained with the film just taken out obtaining them for montmorillonite. For example, it was of the desiccator. For the following ;bservations the practically impossible to make iVa-vermiculite films suitable temperature is increased up t o about 70-90 . 45 minutes :ire for infrared st,udy. allowed t o reach equilibrium. All experiments ~ v c r orepeated t x n or three times. . Other measiirements are odone a t about 150-170°, 220( 5 ) G . Duykaerts, T h c A i i a l y e t , 84, LO1 (1959). 230", 290-300" sild 350-370 . AIont~rnorilloiiitc:
1
sit,
,v(>,, ,vCu0 1 2 + - , i r g o . b 4 2 1
+
I\
)
m
( j 4~
~
~ is
14
12
10 h
2‘
5
u LJ 0
21
14
12
200
400
“C. Fig. 2.--Variation of relative absorbance of the angular vibration band of water with the dehydration temperature. A, montmorillonites: 0 , lithium; 0 , sodium: 0 potassium; 0 , strontium. B: vermiculites: 0,lithium; 0, potassium; 0 , strontium.
11. Experimental Results Figure 2 shows the variation of relative absorbance a t 6.1 p , plotted against dehydration temperature. By “relative absorbance” is meant the ratio of absorbance measured at, a given temperature to the absorbance before heating. This ratio is calculated in order to make the results comparable. The study of these data shows a very important fact: a t temperatures up to 400’) free water still exists in all the samples. For moiitmorillonites, experimental points rcferring to different homoionic clays fall on one smooth curve, while for vermiculites, the variability is greater; moreover, for the lithium sample, relative absorbance follows a distinctly different curve. Figure 3 shows the change of dOOl spacing during dehydration; the average curves are given. Weak values for vermiculites are observed; they are accounted for by the fact that, before the first recording, films were dried under vacuum a t room temperature. It is well known that vermiculite is particularly sensitive to dehydration. Glaeser and PezeratGstudied for us vermiculites (6) W e thank Dr. J. Mering and his co-woihers. It. Claeser and J. I’eaeiat for help and nunicrous suggestions.
10
200
100
C. Fig. 3.-Averagevchange of dOO6 distance with the dehydration temperature. A from above: --, Li-montmorilidnite; --, Sr-montmorillonite, - - Ss-morit,morillonite; ---, K-montmorillonite. B froin above: -, Sr-vermiculite; - -, Xa-vermiculite; ---, K-vermiculite, --, Livermiculite.
which had been-dried under yarious conditions and in suspension. We suminarize their results in Table 11.
Exciiangeable cation
Li Na
I< 8r
llgdratod a t
p / P 0 = 0.59
Air dried
11.95 14.4 10.1 14.4
11.2 12.2
12.7 14.8
11.4 12.8
Suspension
15.0 14.8
12.4 12.4 12.8
..
..
..
..
12.3
14.9
12.5
15.0
..
For all samples, except the potassium vermiculite, the presence of two peaks may be observed, the second one being generally broader than the first. As soon as dehydration proceeds, the first peak disappears. The variation of location of this reflection is given in Fig. 3. It should be pointed out that, in all spectra, the presence of a peak a t 7.2 1;x.subsists after heating half an hour a t 550°, and even after three hours a t GOO0, inasmuch as no reflection occurs at 14.4 kX.
Sept., 1360
D E H Y D R A T I O N OF h f O N T & l O R I L L O K I T E AND V E R M I C U L I T E
From this, the presence of kaolinite may be ruled out, according to prevailing ideas.7 On the other hand, when studying crystals by electron microscopy, numerous curled fla kes arc seen besides thin micaceous plates. As a mat>ter of fact, the question could be raised whether the two high order reflections quoted in Table 11, are not due to the existence of both morphologic features, the tubular form being less expanding than the flat one. During dehydration one reflection only, characteristic of interstratification, can be observed. Comparison between results obtained from montmorillonites and verniiculites shows the different behavior of the clays according to the nature of the base saturation. The order in which collapsing of the layers is observed is for montmorillonites I
