P. 1


|Views: 0|Likes:
Published by Santosh Kumar

More info:

Published by: Santosh Kumar on Apr 14, 2012
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





This article was downloaded by: [Joh Gutenberg Universitaet] On: 12 April 2012, At: 05:46 Publisher: Taylor & Francis Informa Ltd Registered

in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Molecular Physics: An International Journal at the Interface Between Chemistry and Physics
Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tmph20

Equation of state of a hard-core fluid with a Yukawa tail
Douglas Henderson , Eduardo Waisman Lesser Blum
a b e a b f

, Joel L. Lebowitz

c g d


IBM Research Laboratory, San Jose, California, 95193, U.S.A.

Department of Chemistry, University of Illinois, Urbana, Illinois, 61801, U.S.A.

Service de Physique Théorique, Centre d'Études Nucléaires de Saclay, 91190, Gif-sur-Yvette, France

Department of Mathematics, Rutgers University, New Brunswick, N.J., 08903, U.S.A.

Department of Physics, University of Puerto Rico, Rico Piedras, Puerto Rico, 00931

System Science and Software Inc., PO Box 1620, La Jolla, California, 92038

Belfer Graduate School of Science, Yeshiva University, New York, N.Y., 10033, U.S.A. Available online: 22 Aug 2006

To cite this article: Douglas Henderson, Eduardo Waisman, Joel L. Lebowitz & Lesser Blum (1978): Equation of state of a hard-core fluid with a Yukawa tail, Molecular Physics: An International Journal at the Interface Between Chemistry and Physics, 35:1, 241-255 To link to this article: http://dx.doi.org/10.1080/00268977800100181

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The

or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. claims. proceedings. actions.publisher shall not be liable for any loss. Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 . demand.

J. b y g(r)=h(r)+ 1.. T h e total correlation function is related to the radial distribution function ( R D F ) .. 08903. etc.Y. Puerto Rico 00931 (Received 17 ffune 1977) We present Monte Carlo results for the equation of state of a fluid in which the intermolecular potential consists of a hard core of diameter a and an attractive tail which is of the form of a Yukawa function. PRF 8429 AC6 and AFOSR Grant No. g(r) [the probability density. 241-255 E q u a t i o n o f s t a t e o f a h a r d . EDUARDO WAISMANt~ D e p a r t m e n t of Chemistry. the exponential and linearized exponential modifications of the MSA and the generalized MSA all give good results for this system. Rutgers University.S. the MSA is considerably less satisfactory.A. New York. U. LEBOWITZtw Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 Service de Physique Th6orique. and three related approximation schemes. and r i j = Ir i . 1. M. 91190 Gif-sur-Yvette. 10033. Permanent address: System Science and Software Inc. T h e functions h(r) and c(r) are the total and direct correlation functions. wPermanent address: Belfer Graduate School of Science. These results are compared with those obtained from perturbation theory. France and L E S S E R B L U M ~ D e p a r t m e n t of Physics. 1977 : Department of Mathematics. U. N. PO Box 1620.S. the mean spherical approximation (MSA).c o r e f l u i d w i t h a Y u k a w a tail by D O U G L A S HENDERSON I B M Research L a b o r a t o r y . After September. Rico Piedras. U. VOL. 1. California 95193.~ exp {-A(r-a)}/r with A= l'8/a.P.S. Urbana. California 92038.A. 1Q . JOEL L. 35. Yeshiva University.MOLECULAR PHYSICS. La Jolla. N. INTRODUCTION The Ornstein-Zernike (OZ) equation for a simple fluid has the f o r m hl~ = c12+ P S h13Ca~dr3.S. U. New Brunswick. .1978. For a fluid with a hard core h(r)-. Centre d'l~tudes Nucl6aires de Saclay. San Jose. N o . University of Illinois.A. normalized to unity at large separations. 73-2430B at Yeshiva University. While perturbation theory works rather well for this system. However..1 inside the h a r d core. University of Puerto Rico. of finding two molecules separated b y a distance r]. t Supported in part by a grant from the Petroleum Research Foundation No. Illinois 61801.r j l is the scalar distance between the centres of mass of molecules i and j at position r~ and rj. (1) where hlz=h(rl~).A.

and T is the temperature) together with the exact condition h ( r ) = .o)]/r. Henderson et al. They referred to this as the generalized mean spherical approximation (GMSA). for some range of values of A. has been the object of considerable investigation. h(r) = .242 D. Henderson et al. equation (2) can be regarded as the assumed functional form of c(r) for a fluid with some specified intermolecular potential. Because of these applications. the mean spherical approximation (MSA) of Lebowitz and Percus [3] consists in assuming c(r) = -/3w(r). Hoye and Stell [8] have simplified Waisman's solution. r>o.z ( r . r< 1 (2) r > ~. the Hoye-Stell procedure can be used to compute K and z. c(r) = K exp [ . Waisman [2] has obtained an analytic solution of equation (1) for the case where c(r) has a ' Yukawa tail '. . [4] in their work on electrolytes.1. and thus c(r) and h(r). which can also be defined graphically [1]. [9]. was formalized by Hoye et al. While equation (1) is really nothing more than a definition of c(r).1. Their analysis is particularly useful in the GMSA. It has recently been used by Henderson and Blum [5] to give a good description of the properties of hard spheres and by Waisman et al. First. In addition. Waisman's solution can be regarded as the solution of the MSA for a fluid for which w(r) = .or)] rio . directly without iteration. This type of procedure. Waisman's solution.A(r. i. [7] have given series expansions in powers of K and p for the solution.e. r>~ (4) (where/3 = 1/kT. such a fluid could be. its usefulness arises from the fact that in many cases one can obtain a good approximation to g(r) by making a simple ansatz for c(r) and solving (1). Hoye and Stell [10] and Hoye and Blum [11] have generalized Waisman's solution to the case where c(r) is given by a linear combination of Yukawa functions. J where cr is the diameter of the hard core of the molecules. Although there is no real fluid with this intermolecular potential. This solution is of interest in two contexts. [6] to describe accurately hard spheres near a hard wall. (5) where the identifications K=/3e and z = A are made. k is Boltzmann's constant. r> . Waisman et al. Secondly. Thus.~ exp [ . If y =g(e+) and a =/3(OP/OP)T (where p is the pressure and p = N / g is the density) are known. r<cr. a qualitative model of a real fluid. first used by Waisman [1] for hard spheres. The parameters K and z are then not given a priori in terms of the intermolecular potential b u t are determined by some semi-empirical method. consider a fluid with a pairwise additive potential energy for which the intermolecular potential is Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 u(r) = (3) {w(r). which is given in terms of a complex set of simultaneous non-linear equations.j For such a fluid.

e.634 2-561 2"821 2"916 3-127 3"971 4"160 4"227 4"365 4"556 0"6 0"8 .963 2. However. T h i s paper is a necessary step in our study of the interracial regions of this fluid.305 2. In the following sections we then report results obtained from perturbation theory.202 2. T h e uncertainty in p V / N k T and U~/N~ is approximately 0.943 2"561 2"921 2"966 3"205 3"971 4"109 4"257 4"490 4"622 Pert 1-811 2-156 2. T h e M C values of p V / N k T and U~/N~ are given in tables 2 and 3. essentially the method used by Barker and Henderson [15-17] in their study of hard spheres.811 2. In the next section we report results of some M o n t e Carlo computations. Values of g(a+) for the Yukawa fluid (~= 1.176 2. i. the intermolecular potential is given by (3) and (5).040 2"222 2"460 2"561 2"598 2"681 3. respectively. [14] in their study of the square-well fluid which was. 13].811 2.00 1"50 1"00 ~ 2"00 1"50 1"00 0"70 MC 1. The resulting values of g(r) are not given here because of space limitations.129 3-971 4"143 4"204 4"331 4"509 1Q2 p* 0. Table 1.811 2.e. T h e properties of this model Yukawa fluid are determined by several different methods. in turn.8/a).596 2'561 2"806 2"900 3.104 3"971 4.768 1.581 3-629 3"646 3"681 3"729 LEXP 1.in table l we do give our values for g(~+). 12. the M S A and G M S A . Our interest in this Yukawa fluid is due to the fact that expressions for the density profile of this fluid in the simple external potential corresponding to an idealized wall can be obtained [6. Elk ~ 120 K cr ~ 3"4 A). this system permits a direct comparison of the M S A for a simple fluid with other approximations and with exact (Monte Carlo) results. T h e s e are the ' e x p e r i m e n t a l ' results. 2.825 2. In addition.128 2"378 2.166 2.Equation of state 243 Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 In this paper we study the t h e r m o d y n a m i c properties and radial distribution function of a fluid with a hard core plus a Yukawa tail. It should thus be useful in evaluating the strengths and weakness of the M S A and G M S A in treatments of other fluids.330 2. the Yukawa fluid is qualitatively similar to argon for the densities and temperatures of the liquid in equilibrium with its vapour and with about the same potential parameters as the L e n n a r d Jones potential (i.717 2. The radial distribution function.292 2.00 1.811 2. EXP and GMSA 1. All calculations reported here use this value of A. the pressure and the internal energy were calculated by averaging over the chain of configurations.05 and 0-005.141 4"203 4-334 4-517 MSA 1. MONTECARLOC L U A I N A C L TO S The method employed is identical to that used by Henderson et al.561 2"834 2"912 3"140 3"971 4"165 4"235 4"384 4"600 Equation (21) 1. We have f o u n d that with ~ = l'8/cr.561 2-815 2"912 3.4 T* oo 2.376 2.50 1"00 co 2.

p* 0"4 T* ~ 2-00 1"50 1"00 oo 2.5.607 -5.2"622 .00 1.665 .6 0.630 -5.598 -5.672 -2'495 -2'583 .5-608 -5-611 -5.026 -4.658 -2"495 -2"552 .137 -1"668 0"6 Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 4"091 1"594 0"760 -0"911 7"001 3.995 -4.039 .l ' 8 / a ) .3.056 -4.122 0"666 -0'229 4"283 1"978 1"219 -0"283 7"750 4-433 3"332 1.651 -5.969 1"195 -0'360 7'744 4-443 3'340 1'129 -1"722 Et 2"518 1"114 0'657 -0"228 4'283 1"978 1"222 -0"259 7'750 4"446 3"355 1"190 -1"544 EXP P~ 2"518 1"121 0'660 -0"258 4"283 1..619 0.629 -5.006 -4. Values of D.5.476 2"301 -0'049 -3"072 4"283 1"992 1"235 -0"288 7-750 4"464 3"373 1"198 -1"594 0-8 7"744 4"451 3"354 1"160 -1-657 J.8/0).5.5.610 .037 -4"063 -4. (6) go(r) is t h e r a d i a l d i s t r i b u t i o n o f t h e S y s t e m w h e n / 3 c = 0 (i.594 -5.2.017 -4.3.572 .2"638 .635 -5.699 GMSA -2"495 -2"595 .50 1.985 1"224 -0"289 GMSA~ 2'518 1"122 0"655 p* 0'4 T* oo 2"00 1"50 1"00 oo 2"00 1"50 1"00 ov 2"00 1"50 1"00 0"70 MC 2"52 1"08 0"69 -0"21 4'22 2"04 1"21 -0"27 7"65 4"27 3"31 1"29 -1"63 Pert 2"518 1"123 0"664 0"246 4"283 1"985 1.5.2"733 .058 -4.5-611 -5.573 . h a r d s p h e r e s ) . pV/NkT MSA Et for t h e Yukawa fluid ( A .Calculated f r o m e n e r g y equation.3.00 ~ 2.562 P~ 2"518 1"093 0"609 -0"388 4'283 1.3. Calculated f r o m p r e s s u r e equation.594 .145 .00 0.623 -5. Henderson et al.975 -4.030 -4-051 -4.495 -2"586 .125 .573 .2.568 . T a b l e 3.605 -5.626 .035 -4.5.446 3"353 1"185 -1.602 .692 EXP -2"495 -2-592 .975 -4"031 -4.610 -5.226 -0"281 7-750 4'459 3"368 1"195 -1"582 2"518 1.114 . -4. LEXP P$ 2"481 0"943 0"422 0"645 Et 2"518 1"113 0"653 -0"244 4'283 1"977 1"219 -0'271 7"750 4.624 LEXP -2.5. MSA -2"513 -2. PERTURBATION THEORY CALCULATIONS O n e m e t h o d w h i c h h a s b e e n a p p l i e d w i t h g r e a t s u c c e s s i n t h e t h e o r y of d e n s e f l u i d s is t o e x p a n d g ( r ) i n a p e r t u r b a t i o n s e r i e s : g(r) where =go(r) + ~Egl(r ) +.00 1.244 T a b l e 2.3.2"766 .573 .5-589 -5.573 .073 ..573 .5-622 -5.2.2"658 .2"832 .017 -4.050 .e.8 3.50 1.975 -4.625 -5.616 -5.70 Values of MC U~IN~ for Pert the Yukawa fluid ( A = 1..5.975 -4.655 -5.

0245 0.0426 -0"0628 -0"0780 -0-0737 -0.0390 0.0045 0.Equation o[ state 245 w h i c h has a l r e a d y b e e n t a b u l a t e d [15.1417 0"0369 -0"0353 -0"0449 -0"0839 -0"0744 -0.0365 -0"0557 -0"0587 -0"0611 -0"0640 -0"0346 -0"0304 -0"0203 -0"0224 -0"0225 -0"0026 -0-0043 -0. T h e p e r t u r b a t i o n e x p a n s i o n of t h e t h e r m o d y n a m i c p r o p e r t i e s c a n b e o b t a i n e d from A / N k T = A o / N k Y + / 3 ~ A 1 / N k T + (/%)~ & / N k Y + . w h i c h are e x a c t a p a r t f r o m t h e s t a t i s t i c a l u n c e r t a i n t i e s of t h e M C m e t h o d .0265 0"0481 0"0399 0-0203 0"0313 0"0371 0"0152 0"0280 0.0555 -0"0459 -0'0447 -0..4916 1"5379 1"5827 1"6263 1"6688 1"7103 1"7507 1-7903 1-8289 1"8668 1"9039 1"9404 1"976t 2"0112 2"0457 2"0797 2"1131 2"1459 2-1783 2"2102 2"2417 0"2000 0-7809 0-6284 0"4162 0"3111 0"2098 0'1193 0"0815 0"0682 0"0127 0-0186 0"0081 0"0435 0-0040 -0"0212 -0"0534 -0"0131 -0"0105 -0"0091 -0-0241 -0"0126 0"0130 0"0221 0"0149 0"0261 0"0022 0"0020 0"0240 0"0236 .0041 -0. _ l .0168 0"0132 0"0043 0-0025 0"0025 .5000 0"5113 0"3253 0. which is taken to be (7) Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 A o / N k T .0342 0"0433 0"0446 0-0367 0"0526 0"0473 0"0486 0"0354 0"0443 0.0166 0.0"0244 0"0335 0"3000 0"6834 0"4509 0"3312 0"1962 0'1197 0"0544 -0"0538 -0"0326 -0.0116 0.r3rfl) ideal gas t e r m s .0238 0.0051 0-0363 0.o r d e r t e r m .0196 0.0094 0"0042 0"0124 0.0118 0. 16].0827 -0"0781 -0"0515 -0"0563 -0'0389 -0"0277 -0"0123 0"0118 0'0102 0"0199 0"0175 0.4 . T h e f i r s t .0046 0"0133 0"0310 0.0699 -0.0314 -0"0119 -0"0151 -0-0139 -0.0016 0.2903 1"3435 1"3946 1.0154 0"4338 0"3735 0-2483 0-1508 0"0845 0"0287 0"0018 -0"0253 -0"0552 ..0168 -0.2636 0"0266 -0'0075 -0"0027 -0"0147 -0"0192 -0"0122 -0"0138 -0'0110 -0-0102 -0"0060 -0"0038 0. V a l u e s f o r g l ( r ) f o r a h i g h d e n s i t y a r e p l o t t e d in f i g u r e 1.0007 0"0016 0"0034 0"0100 0"0193 0"0184 0"0182 0"0184 0"0089 0"0176 0"0176 0"0227 0"0131 0"0016 0.0110 -0'0082 0. gl(r).0119 0.4440 1. can b e c a l c u l a t e d f r o m t h e c o n v e n i e n t e x p r e s s i o n s ( e q u a t i o n s (59) a n d (66) of r e f e r e n c e [16]) of B a r k e r a n d H e n d e r s o n .0 ' 0 4 8 4 .0061 0"0046 0"0079 r*/p* 1"0000 1"0512 1"1158 1"1769 1"2349 1.0 ' 0 1 7 9 -0"0219 .0363 -0. T h e r e s u l t i n g v a l u e s of gl(r). are l i s t e d in t a b l e 4.0.0243 0"0255 0.. where A 0 is the hard-sphere free energy.0005 0-0003 -0"0013 0"0050 0.+ e (1 (8) Table 4.0 ' 0 7 3 9 -0"0557 -0"0609 -0"0633 -0"0760 -0"0550 -0"0573 -0"0145 -0"0622 -0-0262 -0"0401 . Values of gl(r)for the Yukawa fluid ( ~ = 1-8/~).0051 0. 0"4000 0"5939 0"4194 0"2266 0"1307 0"0191 -0.0175 -0.0077 0"0107 0"0056 0"0166 0"0027 0"0084 0"0006 .0256 0'6000 0"7000 0"8000 0"3178 0-0733 -0"0138 -0"0244 -0"0272 -0"0156 -0"0199 -0"0216 -0"0210 -0.0171 -0.9000 0.0 ' 0 0 9 3 -0"0346 0"0064 -0"0139 0"0024 -0"0019 0'0058 0"0215 0"0118 0"0119 0'0057 0"0198 0"0203 0"0249 0"0204 0"0346 0"0308 0"0352 0"0361 0"0468 0"0417 0"0242 0"0354 0'0451 0"0294 0"0321 0-0307 0"0408 0"0107 0'0178 0.

A x and A 2. Since the gn(a +) are difficult to calculate by direct simulation. For example. the nth-order term in a perturbation expansion of the pressure is obtained from A n by density differentiation. T h e perturbation terms.0'1155351900 T h e An. this pressure equation route is unattractive for . T h e resulting values have been fitted to the function (10) An/Nk T= Cn{1 . 0.2 I F I I ~/ p*=0.8 EXP& LEXP 0. H e n d e r s o n et al. the p~ can also be calculated from the pressure equation. T h e remaining coefficients are chosen by a least-squares criterion. for n = 1 and 2. T h e other thermodynamic properties can be obtained from equation (7) by differentiation.1 Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 1.2 1. T h e values of these coefficients for the Yukawa fluid are given in table 5. Therefore.3 -~ 0. the coefficients are given to high accuracy. However.246 D.4 r/o 1. T h e points give the perturbation theory results which are calculated by MC techniques from exact expressions.1'1397603697 .anP*/(fi~ . If this route is taken. the p~ depend on gn(a+). T h e above expressions result from an integration of the Carnahan and Starling [18] hard-sphere equation of state.5"3886151999 -0"8639815240 . calculated f r o m equation (9).exp [ . are functionals of gn_l(r). (9) where w*(r)=w(r)/e. The curves give the results of MSA and its generalizations. T h e final values of A 1 and A s are residues resulting from cancellation among the terms in (10).0 1. Pn.P*)] . Values for gl(r) for the Yukawa fluid ()~= l'8/a) at a high density. n Coefficients in fit of A1 and A2for the Yukawa fluid (~= 1.4 0.a~P*/fl~} + PnP* + Qn(P*) ~. and thus the p~.8 Figure 1. are obtained f r o m An 1 -N-k-T+~n o I w*(r)g~_l(r) dr. We choose fin = ~/2 and force Pn to give the correct contribution in order p. Table 5.8/a) =n Ca Pn On 1 2 4"75 11"25 0'3150374415 -0"1401430397 . where p * = paa.6 1. Values of A~/NkT are plotted in figure 2.

~ is obtained by differentiating A s.0 Figure 2.08 Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 0.f .9 0"10 0"07 0.6 0.6 0. However. g~(e ) = ~ .. / /" /// EXP & LEXP 0.2 0. the procedure may be inverted to give a method of calculation of the g~(e+). Table 6.02 mA.05 0"03 The perturbation theory value of g(cr+)..7 0"8 0.00 0. . We have not distinguished whether the perturbation theory values of pV/NkT come from the pressure equation or the energy equation since." . respectively.06 ~. Thus. The points give the perturbation theory results which are calculated by MC techniques from exact expressions.8 1./ // / j" / 0.0 p* 0.. calculation of thep~.04 \ \ . if equation (12)is used for g(~+) and equation (6) is used for g(r). the two equations . Values of A2 for the Yukawa fluid (A= 1"8/o). p• g~(a+) p* g~(a+) 0"2 0"3 0-4 0"5 0"34 0"25 0"19 0"14 0. n=O (12) are compared with the M C values in table 1 and the perturbation theory values of pV/NkT and Ui/N~ are Compared with M C values in tables 2 and 3. calculated from 2 g(cr+)= y~ (~)ngn(~+).Equation of state 247 0.4 / // . The solid curve is a least-squares fit of the MC points and the broken curves give the results of the MSA and its generalizations." < 0.) + /2rrp *\-1 (pnV/NkT)+ ~ (l+Zx) exp[-A(x-1)lgn-l(x)xdx' (11) 1 T h e resulting values of gl(e +) where p. and g~(~+) are given in tables 4 and 6. Values of g~(o+) for the Yukawa fluid (A= 1"8/o).">*/ // / 0.

give identical results. Radial distribution function of the Yukawa fluid (A = 1"8/o). w e g i v e t h e p e r t u r b a t i o n t h e o r y v a l u e s of t h e f r e e e n e r g y in t a b l e 7.0. 1.150 .403 0"588 .1.8 ~ = 1..2. "\~ Ms2". since the perturbation theory e q u a t i o n of s t a t e is s e e n to b e g o o d i n t a b l e 2 a n d f i g u r e 4..1"422 2"042 0"048 . \ .0.139 .130 .~.0.039 .0.. T h e points give the M C results and the curves give the results of various approximate theories.631 .00 oo 2"00 1"50 1"00 oo 2"00 1-50 1"00 0"70 Pert 1"130 .05 J r/e I 1.130 .600 .246 4.0-318 .588 -. p e r t u r b a t i o n t h e o r y v a l u e s of g(r) a n d p V / N k T a r e p l o t t e d in f i g u r e s 3 a n d 4.334 . EXP Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 L .024 3.0"590 . " ~ Pert/v\~.653 .025 . 0* 0-4 LEXP 1.4-586 MSA 1.0.0"620 .0"132 . T* oo 2"00 1"50 1.10 ..2. it is r e a s o n a b l e t o b e l i e v e t h a t t h e p e r t u r b a t i o n t h e o r y v a l u e s f o r A a r e a c c u r a t e .1 "485 2"042 0-026 .. T h e r e are no MC values for comparison.. "..130 .2.~. I n a s m u c h as o u r m a i n i n t e r e s t is i n t h e i n t e r r a c i a l r e g i o n s of t h e Y u k a w a f l u i d . ~.2"027 3"403 0.~x.1 "962 3"403 0"613 .0"354 .0"654 .042 0.o .569 ..0"151 .2"247 .. Values of A/NkT for the Yukawa fluid (A = 1.692 EXP 1.x~ !~. However.0"559 .0"355 .1 "976 3.8/0) (perfect gas terms not included). Henderson [ [ et al.859 ....2"183 .248 D.1"493 2"042 0.403 0.0.4"695 0"6 0-8 .0. T h e a g r e e m e n t w i t h t h e m a c h i n e s i m u l a t i o n s is g o o d . I n a d d i t i o n .0 I 1..15 F i g u r e 3. GMSA p*= 0.e s p e c i a l l y at high densities where the perturbation series converges rapidly.206 4-164 T a b l e 7.443 2.

Recently.8335 0"0065 0. Density and pressure of the Yukawa fluid (h=l. the vapour pressure can be calculated easily at low temperatures where the vapour has a vanishingly small density. the free energy of the liquid A L will be roughly constant in the neighbourhood of PL. Once the free energy is known.2 0. need not be known exactly before the vapour pressure is calculated.0 0. Adams [19] has made computer simulation studies of the vapour pressure of the Lennard-Jones fluid and has found that perturbation theory gives accurate values of the vapour pressure of a fluid near its triple point.8.6 0. lnp* = i n T * . f ~ I p I 0. PL. We expect that these values are quite reliable. The situation should be even more favourable here because 3 2 is smaller relative to A 1 than is the case for the Lennard-Jones fluid. Assuming the vapour to be ideal. Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 / 2 I .8/a) in equilibrium with its vapour at T * = 0 " 8 . Table 8. Since the vapour pressure is small. . ICalculated f r o m pressure equation.4 p~ 0.1 +AL/NkT .8375 ~ 0"006 0. T h e points are the M C values and the curves gives the perturbation theory results.8 Figure 4.8336 N 0"006 a Calculated f r o m energy equation. Equation of state of the Yukawa fluid ( t = 1.8/a).8327 0"0059 0. (13) where p*=paS/E and T*=kT/E. The density of the liquid. Presumably this means that perturbation theory is even more rapidly convergent for the Yukawa fluid (with A= l'8/a) than for the Lennard-Jones fluid.Equation of state 249 z2 > Q. Pert Exp(E)t Exp(P)~ GMSA(P)J~ pL* p* 0. The perturbation theory values of PL* and p* are given in table 8 for T*=0.

However. can also be determined.8 for all values of T* and p* which are of interest. v. There are no machine simulation values for comparison. Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 Finally. b. the Fourier transform. and y=g(a +) needed for the calculation of the MSA c(r) can be obtained by iteration from Waisman's equations. they appear as equations (6)-(8) and (10)-(12) in reference [7].Calculated from compressibility equation.g(k). Because of the OZ equation h(k) g(k) 1 . v. there are undoubtedly some errors in these perturbation theory estimates. and y have been determined. can be determined analytically. MEAN SPHERICAL APPROXIMATION As mentioned previously. However. c(r) can be determined. the hard-sphere or infinite temperature result.2 U U N k T . v ~ . is transformed numerically. again by iteration. [7] and have found nearly perfect agreement. and y. b. the MSA for the Yukawa fluid is obtained by combining equations (4) and (5).28 0"141 0"39-0"40 PeVe/NkTe ]. D.g ( k ) and /~0(k)-g0(k). When convergent. (It is to be noted. Evidently. Henderson et al. the four parameters a=fi~p/~p. The contact value is much too low.148 0"38 P~ 1"29 0"27-0. This is seen further in table 1. Because h(r)-c(r) is continuous. We have compared the exact values of these parameters with those obtained from the inverse temperature expansions given by Henderson et al. for completeness the perturbation theory estimates of the critical constants of the Yukawa fluid are given in table 9.250 Table 9. all that remains in t h e determination of g(r)= 1 +h(r) is the numerical transformation of h(k).pg(k)" (14) Thus. a. Once a. five terms in a temperature expansion are sufficient for convergence in the MSA for A= 1. g(k). The resulting values of g(r) at fie= 1 and p*=0-8 are plotted in figure 3. . that a rather obvious additive factor of unity is missing from the second half of equation (8) of reference [7]. the expansion is the easiest method of calculating these parameters. b. this can be done very accurately by transforming only h(k) . These equations are not written out fully in reference [2J. Calculated from pressure equation. the errors should not be large. Given A and fiE.) These factors. particularly if only the difference between h ( k ) . however. 4. from the equations of Hoye and Stell [8]. Values of critical constants for the Yukawa fluid (A = l'8/a). Since the convergence of perturbation theory is slower at low densities. From this.31 0. GMSA Pert To* pc* Pc* 1"255 0"31 0"148 0"38 Ct 1"255 0.

The M S A values of U o A. If equation (16) is combined with the exact condition h(r)= . (16) where Co(r) is the exact hard-sphere direct correlation function. are listed in table 2. [7] have calculated the M S A values for gl(r). The M S A values of p V / N k T . Both go(r) and gl(r) are too small at contact. In most applications the equation of state obtained by differentiating these energy equation values of the free energy are more accurate than those obtained from either the pressure equation or the compressibility. 22] results for hard spheres.e. as is seen in figure 2. calculated from the pressure equation.2 U J N k T . it is not surprising that the MSA g(r) is also too small at contact. These values of UJNE are listed in table 3.8. The MSA go(r) is identical with the Percus-Yevick [21.Equation of state 251 The thermodynamic properties can be calculated from the energy equation. Henderson et al. the absolute error in the free energy itself is small. The free energy can be obtained from these values of U i by numerical integration (analytical integration if the temperature expansion is used). On the other hand. r < a. In fact the integral which appears in (15) is exactly the integral appearing in the pressure equation. However. from a. The equation of state can also be calculated from the compressibility equation. However. LEXP A DEXP MODIFICATIONSOF THE MSA N Andersen et al. and p which are calculated from the energy equation are in good agreement with the MC and perturbation theory results. Normally this is a cumbersome method of obtaining the equation of state. r>cr. ar e significantly in error. This has not been done because it requires a numerical integration which is inconvenient at high temperatures and which is not possible at temperatures below the critical temperature. This is primarily caused by the values of the M S A g(cr+). i. since the integration and the subsequent differentiation must be done numerically. The MSA values for the free energy. Hoye and Stell [20] have shown that the energy equation of state can be calculated from Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 p Z = P~ NkT NkT +3 P*tg~(cr+)-g~ 2~r p*fi" J 1 ~w* g(x)x ~ dx. The errors in the M S A go(r) do not result in serious errors in the MSA A 1. or equivalently from v = .1 . However. the M S A values of p. Thus. (15) This remarkable result permits the calculation of the MSA energy equation of state with no more difficulty than that required from the pressure equation. [23] have proposed the optimized random-phase approximation (ORPA) which consists in replacing equation (4) by c(r)=co(r)-fiw(r). as is seen from table 2. 5. are tabulated in table 7. The terms in the perturbation expansion of the free energy are related to the gn(r) by equation (9). Their values are plotted in figure 1 for p* = 0. for the particular case of the MSA. obtained in this manner. calculated from (15). a. the small values of the M S A gl(r) do result in significant relative errors in A 2. the OZ equation can be .

It is seen that the small differences in the perturbation and EXP free energies (calculated from the energy equation) do not result in significant differences in p* and p*. The ORPA is similar to the MSA. The EXP results are especially good and are somewhat better than the perturbation theory results. therefore. [24] for the square-well potential show that only the hardsphere part of g(r) is significantly affected by the replacement of (4) by (16). Andersen et al. we have assumed the EXP pressure equation vapour pressure is about the same as the EXP energy equation value (this is equivalent to assuming that the EXP pressure equation free energy is about the same as the EXP energy equation result) and have calculated the corresponding value of p* using the pressure equation. if (18) is used. The value is given in table 8 also. somewhat worse than the M S A result. In other words. On the other hand. Both are improvements over the MSA results. result from compensations in higher-order terms.goMSA(r). . The EXP values of the density and pressure of the Yukawa fluid in equilibrium with its vapour at T * = 0 . g(r) =go(r) + CgL(r). 8 are given in table 8. or. as is seen from figure 2. is plotted in figure 1. The LEXP and EXP approximation results [again calculating using (18)] are given in tables 1-3 and 7 and in figure 3. Indeed the perturbation theory results. The result for g(r) can be written as (17) solved and g(r) can be calculated. gt(r). would improve if a larger value of gl(a +) and A s were used. Henderson et al. g(a+) will still be small. where higher-order terms are neglected. The L E X P and EXP gl(r). (20) where go(r) in equations (19) and (20) is the exact hard-sphere RDF. It must. go(r) times the MSA gl(r). Both approximations yield a first-order term.252 D. There is a significant improvement over the M S A values. Since at high densities the differences in the EXP pressure and energy equations of state are similar in magnitude to the differences between the EXP and perturbation theory equations of state and since these latter differences did not result in significant differences in p*. Although g0(a +) is given correctly in the ORPA. calculated from equation (18). it is surprising to note that the L E X P and EXP A 2 is. In fact the recent calculations of Henderson et al. [23] have proposed the E X P (exponential) approximation g(r) =go(r) exp [(gL(r)] and the L E X P (linearized exponential) approximation (19) g(r)=go(r)El+~L(r)]. where go(r) is the exact hard-sphere RDF. This is so despite the fact that the EXP results for gl@ +) and ~/2 are too large in magnitude. which is go(r) times the ORPA gl(r). (18) Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 This means that the ORPA shares many of the deficiencies of the MSA. ~L(r) may be approximated by the difference of the M S A g(r) and the MSA go(r) : ~L(r) ~ g~SA(r) -.

The advantage of the GMSA over the EXP modification of the MSA is that the analytic expression for c(r) is retained.5 Figure 5. We have seen that the MSA energy equations of state is very accurate. The solid and broken lines give the GMSA and MSA results. we could require that the pressure and energy equations of state be consistent. which are used in the MSA. However. use of equation (15) should result in near self-consistency. If we wishes to compute c(r) in the EXP approximation we would have to compute first g(r) (ultimately from a numerical Fourier transformation) and then compute the EXP c(r) by further numerical Fourier transformation with possible loss of accuracy. perturbation. we must know the equation of state before we can use the Hoye-Stell procedure. . respectively. and MSA calculations reported earlier in this paper. it is awkward to do so because equation (15) is valid only in the MSA.5 #e 1. K = fie and z = A. Thus if we neglect the change in the energy equation of state and the change in the integrals in the pressure and energy equations in passing from the MSA to the GMSA.0 0. this is no problem because as a result of the MC. The values of g(a +) calculated from equation (21) are listed in table 1. However. we obtain g(a+) =go(a+) + 89 [g0MSA(a+)]2}. (21) In where g0(a +) is the exact hard-sphere RDF obtained from MC results. or at least nearly so.0 1. principle. However. =8 / v / 0. equation (21) could be generalized into an iterative procedure whereby the pressure and energy equations of state could be made fully self-consistent. GENERALIZEDMEANSPHERICALAPPROXIMATION 253 Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 A final procedure which we examine here is the GMSA where we assume that c(r) has the form given by equation (2) but do not make the identifications.Equation of state 6. The most convenient procedure is to use the Hoye-Stell equations [8] which give K and z explicitly as functions of a=fi~p/~p and y=g(a+). Temperature variation of K in the MSA and GMSA. They are somewhat inferior to those obtained from the EXP approximation. we now know the Yukawa fluid equation of state rather well. On the other hand. p0 * . To calculate g(a+). Because of this we use the EXP values of g(a +) in this work.

The GMSA critical point results are given in table 9. same meaning as in figure 5. On the other hand.5 Figure 6. . The temperature dependence of K and z is qualitatively similar at other densities. We have not calculated the energy equation of state. the GMSA pressure equation of state is slightly better than the EXP pressure equation of state. On the other hand. for fiE=0. Again we have assumed that the GMSA pressure equation of state value of the free energy is about the same as the perturbation theory or EXP energy equation of state values.8 in figures 5 and 6. Temperature variation of z in the MSA and GMSA. The GMSA results are given in tables 1-3 and in figure 3. K increases in an approximately linear fashion. as this would require extensive calculations and presumably would result in values which are much the same as the perturbation theory. K and z can be determined from the Hoye-Stell procedure. LEXP. The resulting values of K and z have been given.0 1. z decreases towards the MSA values as fie increases. and EXP energy equations of state already reported here. The temperature dependence of K and z are given for p*=0. The GMSA pressure equation results are very similar. This should be a very good approximation.254 D. The GMSA compressibility equation results will be identical to the perturbation theory results because of our procedure for choosing a.0 0.8 O Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 0.5 1. Henderson et al. since the perturbation theory equation of state is analytic. but more convenient. to differentiate the perturbation theory pressures. Since the perturbation theory and MSA energy equation of state results are so similar it is equivalent. The curves have the We can calculate a by differentiating the MSA energy equation of state. Presumably the GMSA energy equation results would also be very similar. I I ' o p*=0. As fie increases. Now that a and y are determined. by Henderson and Blum [5]. The GMSA pressure equation estimate the pressure and density of the Yukawa fluid in equilibrium with its vapour are given in table 8.

SMITH. 1.. Phys.J.D. and STELE. 1972. A.. Molec. Rev.... Molec. L. 1977. and STELE. S.. 19. and WAISMAN.. 105. A. Phys. chem. C. Frisch. 1627. and J.. and STELE.. L. J. ft. and BEUM. 34. G... and HENDERSON. Perturbation theory works well. Phys. Phys. Phys. 48... 439. W. STELE. Molec. 16..D.. Phys. 64... 209. 1971. 23.D. chem. 399.. SMITH.D. ADAMS... Rev. CHANDLER. LEBOWITZ. F. J. 1969. Phys. 1958. 1974. Rev. 587. 5026. Molec... D. 1373...D. F. MADDEN. 1976. 32. G... HovE.D.. R. Lett. 1976. of the Yukawa tail is 1. Y. HovE. G... 1966. A. R. Phys. Phys. K. Benjamin. In addition.. Phys. 67. chem. A. modifications of the M S A which do give satisfactory results are considered. 1977.. WAISMAN. Phys. 40.J. S. J. G... SUMMARY 255 Monte Carlo results for the equation of state of a fluid with a hard core and an attractive Yukawa tail have been given for the case where the decay parameter. 1970. Rev... Of these the G M S A has the great advantage of giving analytic expressions for most quantities of interest. 144. statist.F. HovE. G. and YEVICK. 1977.. BARKER. 32. and STELE.. HOvE. 1964.Equation of state 7.W. 195. Molec.. Phys. S. perturbation theory and the mean spherical approximation are applied to this system. 1976. and BEUM.J.. 1291. 1976. Molec.. 1976. 61. A. 187. G.. chem. J. 1976.. BARKER.. STEEL. 110.E.. and HENDERSON. 251.. 1976. G. S. PERCUS.. L.. mad WEEKS. and BARKER.. Molec.K. G. J. L.. D. 1974. 4247. E. Downloaded by [Joh Gutenberg Universitaet] at 05:46 12 April 2012 [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] ... WAISMAN.. WAISMAN. and PERCUS.. (in the press). HENDERSON.W..J. D. 3253. 25. and STARLING. HENDERSON. The M S A is less satisfactory. S.E. edited by H.. Molec.D.. and FITTS.. 1976.). 439.D. E. Phys.D. 635. and STELE. Phys. ft.. 1964.. Chem. and HENDERSON. G. L. L.J. and STELE. 1976. (W. L.E. CARNAHAN. However. ff. 32. 32. HENDERSON. Phys. J. Phys. and HENDERSON.. Phys.. 51. Lebowitz.. 647. 514.. HENDERSON. chem. Phys. HovE. HENDERSON. J. K. ft.. 32. Molec. LEBOWITZ. J. S.N. statist. J. 11-171. 62.. 1976. Phys.J. p. rood. L. Phys. J. J.J. 21...8/~..J.J.. HovE. K. J. and LEBOWITZ. ABRAHAM. S. 411. 31. G. BARKER. Equilibrium Theory of Classical Fluids.H.. Adv. ANDERSEN. Phys.. HENDERSON. p. Inc. 45. ~.. chem. A. 1973. and TACO. ft. 11-32. Phys... 15.. Equilibrium Theory of Classical Fluids.D. Phys. 1976. Chem. 439.D. REFERENCES PERCUS.. BLUM. chem.J.D.

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->