This article was downloaded by: [RMIT University]

On: 17 August 2015, At: 05:48

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,
London, SW1P 1WG

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

Accurate computations of the rovibrational spectrum

of the He–HF van der Waals complex
a a b b
Jose Luis Cagide Fajin , Berta Fernandez , Aleksandra Mikosz & David Farrelly
a
Faculty of Chemistry, Department of Physical Chemistry , University of Santiago de
Compostela , E-15782 Santiago de Compostela, Spain
b
Department of Chemistry and Biochemistry , Utah State University , Logan, Utah 84322,
USA
Published online: 21 Aug 2006.

To cite this article: Jose Luis Cagide Fajin , Berta Fernandez , Aleksandra Mikosz & David Farrelly (2006) Accurate
computations of the rovibrational spectrum of the He–HF van der Waals complex, Molecular Physics: An International Journal
at the Interface Between Chemistry and Physics, 104:09, 1413-1420, DOI: 10.1080/00268970500480984

To link to this article: http://dx.doi.org/10.1080/00268970500480984

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

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. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Molecular Physics, Vol. 104, No. 9, 10 May 2006, 1413–1420

Accurate computations of the rovibrational spectrum of the

He–HF van der Waals complex
JOSE LUIS CAGIDE FAJINy, BERTA FERNANDEZ*y, ALEKSANDRA MIKOSZz and
DAVID FARRELLYz
yFaculty of Chemistry, Department of Physical Chemistry, University of Santiago de Compostela,
E-15782 Santiago de Compostela, Spain
zDepartment of Chemistry and Biochemistry, Utah State University, Logan, Utah 84322, USA

(Received 28 September 2005; in final form 18 November 2005)

The rovibrational spectrum of the He–HF van der Waals complex is calculated from an
accurate intermolecular potential energy surface. This is obtained by fitting a considerable
number of interaction energies evaluated at the Coupled Cluster Singles and Doubles level
including connected triple corrections and with an augmented correlation consistent polarized
Downloaded by [RMIT University] at 05:48 17 August 2015

valence quintuple zeta basis set extended with a set of 3s3p2d1f1g mid-bond functions. The
basis set was selected after a systematic study carried out at four intermolecular geometries.
The potential is characterized by two linear minima, i.e., He–HF and He–FH, with distances
from the He atom to the HF centre of mass of 3.1662 Å and 2.9989 Å and binding energies
of
43.844 and
26.169 cm
1, respectively. These results are compared to data available
in the literature. An analytic fit to this potential energy surface is presented and used to
compute several low-lying rovibrational energy states using coupled channel methods. These
results are compared with fixed-frame diffusion Monte-Carlo calculations using a method
developed specifically for linear molecules.

Keywords: Rovibrational spectrum; CCSD(T) intermolecular potential energy surface; He–HF van der
Waals complex

1. Introduction In recent years a battery of experimental and

The advent of helium nanodroplet isolation spectro-
scopy (HENDI) and its variants [1, 2] has fuelled interest
in van der Waals complexes consisting of clusters of
4
He atoms seeded either with single molecules or small
molecular complexes, e.g., ammonia-dimer [3]. It has
become clear that the rotational and solvation dynamics
of such systems are profoundly affected by whether
or not the droplet is a superfluid. In particular, the
properties of seeded 4He clusters contrast strongly with
the more classical solvation dynamics observed in
fermionic 3He clusters in that rotationally resolved
spectra, albeit with altered molecular constants, are
routinely obtained. Thus, superfluid 4He droplets
constitute ‘‘quantum solvents’’ which serve as excellent
‘‘gentle matrices’’ in which to perform spectroscopic
studies [2].
*Corresponding author. Email: qfberta@usc.es
theoretical methods has been directed towards van der
Waals complexes involving He and other rare-gas atoms
[1–4]. Prominent examples are rare gas complexes with
hydrogen halides [5]. It is, therefore, natural to consider,
specifically, the rotational structure and dynamics of
complexes consisting of 4He atoms and hydrogen halide
molecules (designated HX where X ¼ F, Cl, Br, I).
Besides making contact with experimental studies, a
longer range goal of this work is to uncover and
understand systematic differences in the rotational
dynamics of 4He droplets doped with the homologous
sequence of molecules HF, HCl, HBr, HI and their
dimers, trimers etc. The two extreme limits of this
sequence represent cases where it is expected that the He
density will in the one case (HF) adiabatically decouple
from the molecular rotation or, in the other (HI),
adiabatically follow the rotor [2–4]. The availability of
accurate potential energy surfaces (PESs) for He–HX
complexes is clearly critical to understanding how

Molecular Physics
ISSN 0026–8976 print/ISSN 1362–3028 online ß 2006 Taylor & Francis
http://www.tandf.co.uk/journals
DOI: 10.1080/00268970500480984
 1414 J. L. C. Fajin et al.
(and if ) such predicted transitions actually take place.
In addition, the effect of the He environment on
hydrogen bonding and H-atom tunneling dynamics
within (HX)n, n ¼ 2, 3, . . . , complexes is a topic of
considerable interest [6, 7] which also necessitates the
availability of accurate PESs.
As an initial step towards such a systematic study, in
this paper an accurate PES is developed for the He–HF
van der Waals complex. This surface is then used to
compute several low-lying rotational energy levels using
a variety of accurate quantum approaches including
a variant of the rigid-body diffusion Monte Carlo
(DMC) method [8]. DMC approaches are among the
few methods which are currently available for studying
droplets which consist of a large number of He
atoms doped with a single, or a few, molecules. Here
a fixed frame diffusion Monte Carlo (FFDMC) [4]
method is developed which is specifically applicable to
Downloaded by [RMIT University] at 05:48 17 August 2015
linear rotors such as HF.
He–HX complexes are characterized by two linear
minima whose relative stability depends on the partic-
ular complex [6, 8, 9]. In particular, for the He–HF
van der Waals complex Rodwell et al. [10] used a hybrid
Hartree Fock SCF plus Damped Dispersion (HFD)
method and a gaussian basis set to obtain the potential
energy surface for He–HF complex. They considered the
variation of the H–F distance by using distances of
1.4827, 1.6078, 1.7328, 1.9180 and 2.1032 a0. The inter-
action energies were fitted to a series of Legendre
polynomials, and the potential surface was found to
have an absolute minimum with an equilibrium distance
between the He atom and the HF centre of mass
(denoted Re in the following) of 3.159 Å and an energy
of
33.119 cm
1.
Later Tennyson et al. [11] used the linear combination
of radial and angular momentum function products
(LC-RAMP) method that derives from the close-
coupling approach in body-fixed coordinates, to obtain
the rovibrational levels of the He–HF complex. These
levels were calculated with the potential of Rodwell et al.
[10] The correlation between the HF vibrational coor-
dinate and the complex modes was estimated to be less
than 0.02 cm
1.
Boughton et al. [12] measured the total differential
scattering cross-section of the He–HF complex at two
collision energies. They used the Coupled States (CS),
IOS and single channel spherical potential methods
to predict the cross section from different potentials.
These potentials are based on modifications of those
proposed by Rodwell et al. [10] and of Barker et al. [13]
They concluded that the potential minimum is char-
acterized by Re ¼ (3.11  0.03) Å and an energy of
(
24.0  1.7) cm
1.
In 1990 Lovejoy et al. [14] recorded the near-infrared
rovibrational spectra of the He–HF and the HeDF
complexes in a slit supersonic expansion. To analyze
the data they use a Lennard-Jones(m,6) potential. The
authors also checked the effect of the potential
anisotropy by using the surfaces of Rodwell et al. [10]
but they got discrepancies with respect to the experi-
mental results. Even after modifying these potentials
in order to get better agreement certain discrepancies
still remained. An equilibrium dissociation energy of
(25.03  0.02) cm
1 was obtained for the ground state
and an equilibrium distance Re of (3.113  0.001) Å. The
ground state dissociation energy D0 was estimated
as (7.14  0.10) cm
1.
Smith and Rabitz [15] studied the effect on the
inelastic cross sections, rate constants and rotational
energy level populations of infinitesimal functional
variations in the rigid rotor He–HF surface. The surface
used was the analytic representation of the rigid rotor
Rodwell et al.’s [10] potential proposed by Tennyson
et al. [11]. They found that these observables were
sensitive to the variation of many Legendre components
of the potential.
In 1993 Slee et al. [16] presented the Iterative Secular
Equation (ISE) method which is based on the secular
equation-perturbation theory approach of Hutson and
Le Roy [17]. This method was used together with
Hutson’s potential surface [9c] to obtain the rovibra-
tional energy levels for He–HF, among other complexes.
These energy levels were close to those obtained
by Lovejoy et al. [14]
Moszynski et al. [18, 19] used symmetry-adapted
perturbation theory (SAPT) and medium size basis
sets to compute the intermolecular potential energy
surfaces and the rovibrational spectra of the HeHF and
HeDF complexes. The calculations were performed
for three different HF distances, 1.6078, 1.7328 and
1.9180 a0. The global minimum is in He–HF linear
configuration, with Re ¼ 6.16 a0 and De ¼
39.68 cm
1.
A local minimum is located in the He–FH configuration,
characterized by a He FH distance of 5.59 a0 and an
energy of
36.13 cm
1. This potential energy surface
predicts a dissociation energy of 7.38 (7.50) cm
1 for
He–HF (He–DF) and compares very well with the
empirical potential of Lovejoy and Nesbitt [14] although
around the T-shaped configuration the two surfaces are
slightly shifted and the agreement is worse. Later
Moszynski et al.’s potential was slightly modified [20]
and used to calculate the integral and differential cross
sections for elastic and inelastic He–HF scattering.
The results were compared with previous experimental
data given accurate scattering cross sections.
In 2003, Stoecklin et al. [21] using the Brueckner
Coupled Cluster Doubles method with a perturbative
 Accurate computations of the rovibrational spectrum
estimation of the triple contribution (BCCD(T)) method
and the aug-cc-pVQZ basis set extended with bond
functions calculated the potential energy surface for the
He–HF complex. The calculations were carried out for
459 geometries with three different H–F distances, 1.50,
1.7328 and 2.12 a0. After fitting the potential, the surface
presented two minima. The global minimum is located
in the linear He–HF conformation, with Re ¼ 5.95 a0
and it has an energy of
43.70 cm
1. The local minimum
is in the He–FH geometry with the He atom at 5.65 a0
from the HF centre of mass and an energy of

25.88 cm
1. The barrier between the two minima
is given by an almost T-shaped complex geometry with
R ¼ 6.10 a0 and an energy of
16.7532 cm
1. The
surface was used to determine the quenching cross
section and rate coefficient of HF molecules in collisions
with He atoms by performing close coupling calcula-
tions. In a subsequent work, Reese et al. [22] used this
Downloaded by [RMIT University] at 05:48 17 August 2015
surface to study the rotational transitions induced
in these collisions.
In previous work, highly accurate intermolecular
potential energy surfaces have been computed using
the coupled cluster singles and doubles model including
connected triple corrections (CCSD(T)) and augmented
correlation consistent basis sets extended with a set of
3s3p2d1f1g mid-bond functions (denoted 33211) [23].
In the present study we evaluate the He–HF van der
Waals complex ground state PES and the corresponding
rovibrational levels. Considering the good results previ-
ously provided by the CCSD(T) method [23], we use
it here. In order to select an adequate basis set, and
knowing the importance of properly checking basis set
convergence (see for example the analysis in Ref. [23b]),
we decide to start by carrying out a systematic basis set
study within the series of bases xaug-cc-pVXZ-33211
(x ¼
,d; X ¼ D, T, Q, 5, 6).
The paper is organized as follows: In section 2
we describe the evaluation of the PES and comment
on the results. In section 3, we compute the rovibra-
tional spectrum using standard methods. The fixed-
frame diffusion Monte-Carlo method is described in
section 4 and ground state energies are computed. The
excited states are obtained using the fixed-node
approach. In the last section, we provide a summary
and give our concluding remarks.
2. Intermolecular potential energy surface
We select a set of 110 intermolecular geometries to carry
out the ab initio interaction energy calculations. The R,

coordinates are use to describe these geometries; R
is the length of the He position vector whose origin
is located at the HF center of mass and
the angle
1415
this vector makes with the HF bond axis, taking as
¼ 0
the linear He–HF configuration. The geometry of the
HF molecule is kept fixed at the r0 value of 0.925597 Å
obtained from the microwave spectra [24].
Interaction energies were calculated using the super-
molecular model and corrected for basis bet superposi-
tion error (BSSE) by invoking the counterpoise method
of Boys and Bernardi [25]. The BSSE correction is
known to be essential in order to get accurate interaction
energies for these systems. By using the CCSD(T)
method and Dunning’s augmented correlation consis-
tent polarized valence basis sets extended with an
additional set of 3s3p2d1f1g mid-bond functions we
obtained very good results in previous studies on similar
complexes (see Ref. [23b] and Ref. cited therein).
Therefore, we use this method and these bases to carry
out the interaction energy calculations. The mid-bond
functions [26] are well known to very efficiently improve
basis set convergence, they are located in the middle of
the van der Waals bond, and denoted 33211. The
exponents of these functions are: 0.90, 0.30, 0.10 for the
s and the p functions; 0.60, 0.20 for the d functions, and
0.30 for the g and f functions.
In the study of NeN2 van der Waals complex [27], we
showed that to obtain accurate van der Waals complex
rovibrational spectra it is essential to carefully select
an adequate basis set. Considering this, we start the
calculations with a systematic basis set convergence
study using the xaug-cc-pVXZ (x ¼
, d; X ¼ D, T, Q, 5)
set of bases, with and without the 33211 set of mid-bond
functions. We carry out the calculations at four geome-
tries: two linear selected close to the PES minima
available in the literature (He–HF: R ¼ 3.26 Å,
¼ 0 ;
He–FH: R ¼ 2.9581 Å,
¼ 180 ) [18], a T-shape geom-
etry (R ¼ 3.6609 Å,
¼ 90 ) and a geometry in the
positive energy region (R ¼ 2.79 Å,
¼ 0 ). The results of
this study are presented in Table 1.
Taking into account the size of the interaction energy,
we consider an error limit of 0.5 cm
1 in the analysis. As
expected, we can see that the presence of the mid-bond
functions accelerates basis set convergence, and even at
the quintuple zeta level these functions are relevant
in the basis set. We have to go up to the augmented
quintuple zeta level in order to get the convergence-limit
stability. The double augmentation of this basis set is
not needed, but in the calculations performed without
mid-bond functions it is. It is worthwhile to point
out the large improvement in the results at the double
zeta level when using mid-bond functions.
The He–HF configuration is significantly more
stable than the He–FH, and all bases agree on this
result, although differing considerably in the relative
energy value. Considering the above, to carry out the
calculations we select the aug-cc-pV5Z-33211 basis set.
 1416 J. L. C. Fajin et al.

Table 1. Basis set study. Energies are given in cm
1. See text We use the DALTON [28] and ACES II [29]
for the definition of the different geometries. programs. The interaction energy results are displayed
BASIS HeHF HeFH T-Shape Positive in table 2.
The complex’s PES is constructed by fitting the
aug-cc-pVDZ-33211
41.399
26.431
12.869 15.434
aug-cc-pVTZ-33211
42.182
25.784
12.156 14.250
ab initio single point interaction energies to the ana-
aug-cc-pVQZ-33211
42.356
25.856
12.024 11.684 lytic function V(R,
), originally suggested by Bukowski
aug-cc-pV5Z-33211
42.467
25.969
12.023 10.566 et al. [30] and that has been used in several similar
daug-cc-pVDZ-33211
40.773
26.602
12.871 16.279 potentials [23g, 31, 32]. The function is the sum of
daug-cc-pVTZ-33211
42.082
25.748
12.202 14.224 two terms, a short range term Vsh and an asymptotic
daug-cc-pVQZ-33211
42.386
25.918
12.056 11.518 term Vas:
daug-cc-pV5Z-33211
42.487
26.013
12.046 10.446
aug-cc-pVDZ
31.335
9.156
7.741 69.152
VðR,
Þ ¼ Vsh ðR,
Þ þ Vas ðR,
Þ ð1Þ
aug-cc-pVTZ
38.598
20.206
10.195 23.574
aug-cc-pVQZ
41.524
23.230
11.148 12.896
aug-cc-pV5Z
42.207
24.633
11.547 11.022 where
daug-cc-pVDZ
33.929
14.422
8.996 63.499
daug-cc-pVTZ
39.372
23.523
11.265 21.683 Vsh ðR,
Þ ¼ GðR,
ÞeDð
Þ
Bð
ÞR ð2Þ
daug-cc-pVQZ
42.083
25.120
11.876 12.431
daug-cc-pV5Z
42.392
25.634
11.957 10.782
Downloaded by [RMIT University] at 05:48 17 August 2015

Table 2. Intermolecular energies in cm
1, R in Å and
in degrees.

R
E R
E R
E
7.0000 0.0
0.387 5.0000 60.0
2.243 2.8500 30.0
9.484
6.0000 0.0
1.053 4.5000 60.0
4.364 2.7500 30.0 20.197
5.5000 0.0
1.878 4.0000 60.0
8.890 2.6000 30.0 111.126
5.0000 0.0
3.567 3.5000 60.0
17.274 2.5000 30.0 224.032
4.5000 0.0
7.278 3.0000 60.0
17.319 6.0000 60.0
0.705
4.2500 0.0
10.664 2.9500 60.0
14.184 5.5000 60.0
1.224
4.0000 0.0
15.789 2.9000 60.0
9.638 2.4000 150.0 171.941
3.8500 0.0
19.977 2.8500 60.0
3.320 2.3000 150.0 311.593
3.7500 0.0
23.300 2.7500 60.0 16.502 2.5000 15.0 282.719
3.6500 0.0
27.044 2.6000 60.0 75.369 2.5000 45.0 172.538
3.5000 0.0
33.269 2.5000 60.0 147.092 2.5000 135.0 109.412
3.2600 0.0
42.468 5.5000 90.0
1.080 2.5000 165.0 67.788
3.2500 0.0
42.738 5.0000 90.0
1.966 5.5000 30.0
1.614
3.1500 0.0
44.017 4.5000 90.0
3.796 5.0000 30.0
3.018
3.0749 0.0
42.591 4.0000 90.0
7.635 4.5000 30.0
6.035
3.0000 0.0
37.820 3.6608 90.0
12.024 4.0000 30.0
12.781
2.9500 0.0
31.962 3.5000 90.0
14.442 3.5000 30.0
26.208
2.9000 0.0
23.200 3.0000 90.0
12.156 3.0000 30.0
29.567
2.8900 0.0
21.029 2.9500 90.0
8.908 2.9500 30.0
25.241
2.8749 0.0
17.476 2.9000 90.0
4.368 2.9000 30.0
18.809
2.8600 0.0
13.656 2.8500 90.0 1.807 3.5000 150.0
15.735
2.8500 0.0
10.696 2.7500 90.0 20.883 3.0000 150.0
22.306
2.8000 0.0 6.598 2.6000 90.0 76.290 2.8500 150.0
15.771
2.7000 0.0 61.081 2.5000 90.0 142.970 2.8000 150.0
10.942
2.5000 180.0 60.674 6.0000 120.0
0.634 2.7500 150.0
4.106
2.6000 180.0 17.117 5.5000 120.0
1.087 2.6500 150.0 18.068
2.7000 180.0
7.059 5.0000 120.0
1.968 2.6000 150.0 35.015
2.8000 180.0
19.449 4.5000 120.0
3.786 2.5000 150.0 86.364
2.9000 180.0
24.833 4.0000 120.0
7.633 5.5000 150.0
1.109
2.9580 180.0
25.970 3.5000 120.0
14.794 5.0000 150.0
1.991
3.0000 180.0
26.189 3.0000 120.0
15.114 4.5000 150.0
3.808
3.0100 180.0
26.181 2.9500 120.0
12.459 4.0000 150.0
7.729
3.2500 180.0
22.231 2.9000 120.0
8.581 5.0000 180.0
1.100
3.5000 180.0
16.163 2.8500 120.0
3.157 5.5000 180.0
1.114
3.7500 180.0
11.226 2.7500 120.0 14.014 6.0000 180.0
0.658
4.0000 180.0
7.741 2.6000 120.0 65.676 31641 30.9
33.773
4.5000 180.0
3.800 2.5000 120.0 129.372
 Accurate computations of the rovibrational spectrum 1417

and Table 3. Parameters of the analytic PES fitted to the ab initio

interaction energies.
X
2
Cl6 0
Vas ðR,
Þ ¼ f6 ðBð
ÞRÞ  P ðcos
Þ ð3Þ Parameter Value Parameter Value
l¼0,2...
R6 l
b0 13.35140347292370 g11
0.48875737093263
Dð
Þ, Bð
Þ, and GðR,
Þ are expansions in Legendre b1
0.09056257250910 g12
0.66202397351030
polynomials (P0l ); b2 0.36222620495196 g13
0.61187512997162
b3 0.10752025627937 g14
0.19932910113773
X
5 b4
0.10729093395687 g15 0.02744036687153
Bð
Þ ¼ bl P0l ðcos
Þ, ð4Þ b5 0.09867415096543 g20 1.04324447891595
l¼0 d0
2.53146116070142 g21 0.09402772497221
d1 0.07421267974884 g22 0.20171778104736
X
5 d2
0.08276071628834 g23 0.16229578225988
Dð
Þ ¼ dl P0l ðcos
Þ, ð5Þ d3
0.01498993894000 g24 0.03785716244561
l¼0 d4 0.04330076241291 g25
0.00646574608438
d5
0.03802078606627 g30
0.10441433253613
X
5
 g00 4.57186377381132 g31
0.00497315571833
GðR,
Þ ¼ g0l þ g1l R þ g2l R2 þ g3l R3 P0l ðcos
Þ ð6Þ g01 0.75361611194433 g32
0.02153063761904
l¼0 g02 0.72139454377371 g33
0.01500658270722
g03 0.78271512913456 g34
0.00145052999200
Downloaded by [RMIT University] at 05:48 17 August 2015

and g04 0.29669542735975 g35 0.0000000000000

g05
0.02516881784584 C06
0.01401238000863
6

x
xk =k!
g10
3.73235058315367 C26 0.00793751597814
f6 ðxÞ ¼ 1
e k¼o ð7Þ

is the Tang-Toennies damping function and x ¼ Bð
ÞR. HeHF contour

5.5
bl, dl, gkl and Cl6 are adjustable parameters. The fitted
values of the corresponding PES parameters are pre-
5
sented in table 3.
−3
The fit of the ab initio values gives a standard error of
0.0106 cm
1 and a maximum error of 0.3235 cm
1. The 4.5
R (Angstrom)

geometry corresponding to this error is R ¼ 6.0 Å and

¼ 0 . A contour plot of the PES is presented in 4
figure 1. Two minima are found in the PES, the absolute
minimum has energy of
43.844 cm
1 and it is located 3.5
at the He–HF linear configuration with the He atom at a −15
−42
distance of 3.1662 Å with respect to the HF center of 3 −24
mass. The surface has a second minimum at the He–FH
linear configuration, with the He at a distance
2.5
of 2.9989 Å with respect to the HF center of mass and 0 20 40 60 80 100 120 140 160 180
at an energy of
26.169 cm
1. A saddle point with T (degrees)
energy of
16.882 cm
1 is located at R ¼ 3.2277 Å and Figure 1. Contour plot in Jacobi coordinates of the

¼ 96.222 . This gives an energy barrier between the aug-cc-pV5Z-33211.
global and the local minima of 26.962 cm
1.
In table 4, these results are compared to those Taking into account the good performance of the
previously available. They are considerably different method and basis set we use, we can consider the dif-
from those in Ref. [10, 12, 14], and the SAPT [18], but ferences with respect to the results in Ref. [10, 12, 14, 18]
very close to those of previous theoretical calculations large, and suggest that these studies need revision.
with the BCCD(T) model [21]. The SAPT surface is
characterized by a difference between the He–HF and
He–FH minima of only 3.55 cm
1, whereas we get a 3. Rovibrational levels and fixed frame diffusion
difference of 17.675 cm
1 (in good agreement with Ref. Monte Carlo calculations
[21]). The SAPT potential differs also considerably from
the experimental results of Ref. [14] (up to 14.65 cm
1 in Using the analytical fit to the potential several low-lying
the absolute minimum). rovibrational energy levels were calculated using the
 1418 J. L. C. Fajin et al.

Table 4. Stationary points. Comparison to previous results.

HeHF HeFH Saddle point

R(Å) E(cm
1) R(Å) E(cm
1) R(Å)
( ) E(cm
1)
This work 3.1662
43.844 2.9989
26.169 3.2277 96.222
16.882
CCSD(T)
HFD[10] 3.159
33.119
Exp. [12] 3.11  0.03a
24.1  1.7
Exp. [14] 3.113  0.001
25.0  0.2
SAPT[18] 3.2597
39.68 2.9581
36.13 3.2539 90.0
19.028
BCCD(T)[21] 3.1486
43.70 2.9899
25.88 3.2280 95.0
16.7532
a
This is an Rm value.

Table 5. Rovibrational energy levels, given in cm
1 and labeled by the set of quantum numbers (J,Ka,Kc).

Level This Work CCSD(T) This Work BOUND LC-RAMP[11] ISE[16] SAPT[19] L-J(m,6)[14]
HeHF (0,0,0)
6.716
6.720
6.393
7.349
7.380
7.347
Downloaded by [RMIT University] at 05:48 17 August 2015

(1,0,1)
5.964
5.968
5.632
6.574
6.608
6.572

(2,0,2)
4.484
4.487
4.128
5.045
5.085
5.043
(3,0,3)
2.326
2.233
1.924
2.814
2.861
2.812
(4,0,4) 0.003
0.040 0.011
HeDF (0,0,0)
7.501
7.517
(1,0,1)
6.731
6.743
(2,0,2)
5.212
5.216
(3,0,3)
2.991
2.984
(4,0,4)
0.173
0.153

TRIATOM program suite [33] and also BOUND [34]
which is a close coupling approach. These calculations
are reported in table 5 and excellent agreement is
obtained between the two methods. In the BOUND
calculations the potential was re-expanded in a basis of
Legendre functions up to order eight. Of course, these
methods are not readily extended to complexes contain-
ing more than a single He atom and other approaches
must be used. In particular, quantum Monte Carlo
methods [8] have been used successfully to study a large
variety of seeded He clusters. DMC methods, often
coupled with importance sampling, have been used
extensively to study the properties of seeded He
droplets [4]. Here we develop a variant of DMC that is
designed to treat linear rotors, or collections of linear
rotors, embedded in He nanodroplets.
In the center-of-mass frame, the Hamiltonian for a
rigid linear rotor such as HF interacting with a He atom
is given by
2 2
h the molecule. This problem can be avoided by recogniz-
H¼
r þ VðR,
Þ þ Hmon
2
where  is the reduced mass and the monomer
Hamiltonian, Hmon, is diagonal in a frame fixed on
the rotor with the z-axis coinciding with the HF
bond, i.e.,
Hmon ¼ BðJ2x þ J2y Þ: ð9Þ
Here, B is the rotational constant. In rigid-body, DMC
(RBDMC) rotations and translations are treated exactly
on the same footing [4, 8], i.e., the components of the
total angular momentum Jx, Jy Jz, are thought of as
generating rotations about the x- and y- and z-axes just as
the momentum operators px, py, pz generate translations
along these axes. Of course doing this ignores the fact
that angular momentum components do not commute
with each other. However, for small enough rotational
moves in the RBDMC method the errors introduced
by this approximation to the Green function are gener-
ally negligible. A complication is apparent however; the
Hamiltonian used in RBDMC is, of necessity, defined in
the laboratory frame (otherwise Coriolis terms arise
which are hard to treat in DMC [4]). However, the rotor
Hamiltonian is diagonal only in a frame that rotates with
ð8Þ ing that the monomer Hamiltonian can be expressed in
terms of the orbital angular momentum operator, i.e.,
Hmon ¼ BL2 ð10Þ
 Accurate computations of the rovibrational spectrum
where L ¼ (Lx,Ly,Lz). In this procedure, rotational
diffusive moves are made around the center-of-mass
axes which are always aligned with the laboratory axes.
The ground state energy obtained using this method is
E ¼
6.718  0.059 cm
1 in excellent agreement with
other methods reported in table 5. An expanded
discussion of this algorithm, including the incorporation
of importance sampling, together with application to
the calculation of ground and excited states of larger
clusters, doped with HF, HCl, HBr and HI molecules
will be described elsewhere.
4. Summary and conclusions
The potential energy surface of the He–HF van der
Waals complex is obtained by fitting 110 interaction
energies obtained with the CCSD(T) method and the
Downloaded by [RMIT University] at 05:48 17 August 2015
aug-cc-pV5Z basis set extended with a set of 3s3p2d1f1g
mid-bond functions. This basis set is selected after
a systematic study carried out at four intermolecular
geometries. The potential has two linear minima, i.e.
He–HF and He–FH, with distances from the He atom
to the HF centre of mass of 3.1662 and 2.9989 Å,
respectively; and with binding energies of
43.844 and

26.169 cm
1, respectively. The low-lying rovibrational
energy levels of the complex are computed using
different methods and good agreement is obtained
among these.
Acknowledgements
This work has been supported by the European
Research and Training Network NANOQUANT,
contract No. MRTN-CT-2003-506842; by the Spanish
Ministerio de Educación y Ciencia and FEDER
(CTQ2005-01076 project); by the Petroleum Research
fund, administered by the American Chemical Society;
and by the US National Science Foundation though
grant 0202185. We acknowledge computer time from
CESGA.
References
[1] J. P. Toennies and A. F. Vilesov, Annu. Rev. Phys. Chem.
49, 1 (1998).
[2] C. Callegari, K. K. Lehmann, R. Schmied, and G. Scoles,
J. Chem. Phys. 115, 10090 (2001).
[3] M. Behrens, U. Back, R. Frochtenicht, M. Hartmann, and
M. Havenith, J. Chem. Phys. 107, 7179 (1997).
[4] E. Lee, D. Farrelly, and K. B. Whaley, Phys. Rev. Lett.
83, 3812 (1999).
[5] K. Nauta and R. E. Miller, J. Chem. Phys 113, 10158
(2000).
1419
[6] A. Sarsa, Z. Bacic, J. W. Moskowitz, and K. E. Schmidt,
Phys. Rev. Lett. 88, 123401 (2002).
[7] H.-C. Chang and W. Klemperer, J. Chem. Phys. 104, 7830
(1996).
[8] V. Buch, J. Chem. Phys. 97, 726 (1992).
[9] (a) J. M. Hutson and B. J. Howard, Mol. Phys. 45, 769
(1982); (b) C. M. Lovejoy and D. J. Nesbitt, Chem. Phys.
Lett. 147, 490 (1988); (c) J. M. Hutson, J. Chem. Phys. 89,
4550 (1988); (d) J. M. Hutson, J. Chem. Phys. 91, 4448
(1989); (e) G. Chalasinski, M. M. Szczesniak, and
B. Kukawska-Tarnawska, J. Chem. Phys. 94, 6677
(1981); (f ) G. Chalasinski, S. M. Cybulski,
M. M. Szczesniak, and S. Scheiner, J. Chem. Phys. 91,
7048 (1989).
[10] W. R. Rodwell, L. T. Sin-Fai-Lam, and R. O. Watts,
Mol. Phys. 44, 225 (1981).
[11] J. Tennyson and B. T. Sutcliffe, J. Chem. Phys. 79, 43
(1983).
[12] C. V. Boughton, R. E. Miller, P. F. Vohralik, and
R. O. Watts, Mol. Phys. 58, 827 (1986).
[13] J. A. Barker and A. Pompe, Aust. J. Chem. 21, 1683
(1968).
[14] C. M. Lovejoy and D. J. Nesbitt, J. Chem. Phys. 93, 5387
(1990).
[15] M. J. Smith and H. Rabitz, Chem. Phys. 150, 361 (1991).
[16] T. Slee and R. J. Le Roy, J. Chem. Phys. 99, 360
(1993).
[17] J. M. Hutson and R. J. Le Roy, J. Chem. Phys. 83, 1197
(1985).
[18] R. Moszynski, P. E. S. Wormer, B. Jeziorski, and
A. van der Avoird, J. Chem. Phys. 101, 2811 (1994).
[19] R. Moszynski, B. Jeziorski, A. van der Avoird, and
P. E. S. Wormer, J. Chem. Phys. 101, 2825 (1994).
[20] R. Moszynski, F. de Weerd, G. C. Groenenboom, and
A. van der Avoird, Chem. Phys. Lett. 263, 107 (1996).
[21] T. Stoecklin, A. Voronin, and J. C. Rayez, Chem. Phys.
294, 117 (2003).
[22] C. Reese, T. Stoecklin, A. Voronin, and J. C. Rayez,
A. & A. 430, 1139 (2005).
[23] (a) C. R. Munteanu, J. L. Cacheiro, B. Fernández, and
J. Makarewicz, J. Chem. Phys. 121, 1390 (2004);
(b) J. L. Fajı́n, J. L. Cacheiro, and B. Fernández,
J. Chem. Phys. 121, 4599 (2004).
[24] D. Webb and K. N. Rao, J. Mol. Spect. 28, 121 (1968).
[25] S. F. Boys and F. Bernardi, Mol. Phys. 19, 553 (1970).
[26] F. M. Tao and Y. K. Pan, J. Chem. Phys. 97, 553
(1992).
[27] C. R. Munteanu, J. L. Cacheiro, and B. Fernández,
J. Chem. Phys. 120, 9104 (2004).
[28] T. Helgaker, H. J. Aa. Jensen, P. Jørgensen, J. Olsen, K.
Ruud, H. Ågren, A. A. Auer, K. L. Bak, V. Bakken,
O. Christiansen, S. Coriani, P. Dahle, E. K. Dalskov,
T. Enevoldsen, B. Fernández, C. Hättig, K. Hald,
A. Halkier, H. Heiberg, H. Hettema, D. Jonsson,
S. Kirpekar, R. Kobayashi, H. Koch, K. V. Mikkelsen,
P. Norman, M. J. Packer, T. B. Pedersen, T. A. Ruden,
A. Sánchez, T. Saue, S. P. A. Sauer, B. Schimmelpfennig,
K. O. Sylvester-Hvid, P. R. Taylor, O. Vahtras, DALTON –
an electronic structure program, release 1.2 (2001).
[29] J. F. Stanton, J. Gauss, J. D. Watts, M. Nooijen, N.
Oliphant, A. S. Perera, P. G. Szalay, W. J. Lauderdale,
S. A. Kucharski, S. R. Gwaltney, S. Beck, A. Balkova,
D. E. Bernholdt, K. K. Baeck, P. Rozyczko, H. Sekino,
C. Hober, R. J. Bartlett, J. Almlöf, P. R. Taylor,
 1420 J. L. C. Fajin et al.
T. Helgaker, H. J. Aa. Jensen, P. Jørgensen, J. Olsen,
ACES II is a program product of the Quantum Theory
Project, University of Florida, 1996, 32611 .
[30] R. Bukowski, J. Sadlej, B. Jeziorski, P. Jankowski,
K. Szalewicz, S. A. Kucharski, H. L. Williams, and
B. M. Rice, J. Chem. Phys. 110, 3785 (1999).
[31] R. R. Toczylowski and S. M. Cybulski, J. Chem. Phys.
112, 4604 (2000).
Downloaded by [RMIT University] at 05:48 17 August 2015
[32] T. B. Pedersen, J. L. Cacheiro, B. Fernández, and
H. Koch, J. Chem. Phys. 117, 6562 (2002).
[33] J. Tennyson, S. Miller, and C. R. Le Sueur, Comput.
Phys. Commun 75, 339 (1993).
[34] J.M. Hutson, BOUND computer code, version 4 (1992),
distributed by the Collaborative Computational
Project No. 6 of the Science and Engineering Research
Council (UK).