Synthesis and resolution of 3,3?-disubstituted xyLBINAP derivatives and their application in rhodium-catalyzed asymmetric hydrogenation

Article abstract of DOI:10.1055/s-0029-1217741

A novel class of 3,3-disubstituted xylBINAP ligands have been synthesized and tested in the hydrogenations of substituted olefins. This new substitution pattern has demonstrated that the 3,5-dialkyl meta effect and 3,3-disubstitution can operate in a syne

Full text of DOI:10.1055/s-0029-1217741

LETTER
2513
Synthesis and Resolution of 3,3¢-Disubstituted xylBINAP Derivatives and
Their Application in Rhodium-Catalyzed Asymmetric Hydrogenation
3
, 3
¢
-Disubstituted
a
xylBIN
A
P
nDerivatives ica A. Rankic, J. Matthew Hopkins, Masood Parvez, Brian A. Keay*
Department of Chemistry, University of Calgary, 2500 University Drive NW, Calgary, AB, T2N 1N4, Canada
Fax +1(403)2899488; E-mail: keay@ucalgary.ca
Received 29 April 2009
ee values as high as 74.7% were achieved by using ligand
4 with OC(O)tBu groups in the 3 and 3¢ positions of the
binaphthalene framework. Although 3,3¢-disubstitution
provided an increase in enantioselectivity for the hydro-
Abstract: A novel class of 3,3¢-disubstituted xylBINAP ligands
have been synthesized and tested in the hydrogenations of substitut-
ed olefins. This new substitution pattern has demonstrated that the
3,5-dialkyl meta effect and 3,3¢-disubstitution can operate in a syn-
ergistic fashion in Rh-catalyzed hydrogenation of dehydroamino genation of methyl N-acetamido cinnamate, we saw the
acids. Notably, (Sax)-8 outperforms BINAP, xylBINAP and previ-
ously reported 3,3¢-disubstituted BINAP derivatives in the hydro-
genation of methyl N-acetamido cinnamate.
opportunity for further improvement.
R
R
Key words: ligands, hydrogenations, asymmetric catalysis, homo-
geneous catalysis, rhodium
PPh2
PPh2
PAr2
PAr2
Asymmetric catalysis is one of the most powerful meth-
ods for forming an array of enantioenriched chiral struc-
tures. Transition-metal-based processes, which make up a
large component of this field, rely on an organic ligand to
impart chirality on the forming product during the catalyt-
ic cycle. Although a large pool of chiral ligands is avail-
R
R
(S)-1 R = H
(R)-6 R = H
(Rax)-7
(Sax)-8
(S)-2 R = Oi-Pra
(S)-3 R = OMe
(S)-4 R = OC(O)t-Bu
(S)-5 R = OBn
O
R =
(R)-9 R = Oi-Pr
OMe
(S)-10 R = OC(O)t-Bu
able, there is still a need for new ligands, as not all of the Figure 1 3,3¢-Disubstituted BINAP and 3,3¢-disubstituted xylBINAP
derivatives
existing ligands are suitable for all substrates. One method
of developing new chiral ligands is to optimize the elec-
tronic and steric properties of an existing ligand through
systematic modification of its framework.
We envisioned that exchanging the phenyl rings in our
3,3¢-disubstituted BINAP derivatives 2–5 for 3,5-dimeth-
ylphenyl (m-xylyl) rings would allow us to take advantage
of the 3,5-dialkyl meta effect.4 It has been demonstrated
that using 3,5-dialkylphenyl substituents on phosphorus
in place of simple phenyl rings, causes the chiral pocket to
become more rigid and well-defined due to hindered rota-
tion about the P–Cipso bonds, which translates into higher
catalyst selectivity in some systems. We were interested
in exploring how increasing the steric congestion in both
the 3 and 3¢ positions of the binaphthyl skeleton, as well
as on the phosphorus atoms, would affect the enantio-
selectivity of Rh-catalyzed hydrogenations of methyl
N-acetamido cinnamic acid as well as other hydrogena-
tion substrates. Herein, we report the synthesis and reso-
lution of ligands 7–10 (Figure 1) and their application in
the hydrogenation of dehydroamino acids and enamides.
Phosphinylation of 4-bromo-2-naphthol5 (11) using
di(3,5-dimethylphenyl)phosphinic chloride6 in the pres-
ence of DMAP and Et3N, afforded phosphinate ester 12
(Scheme 1). Lithiation-induced migration of the di-
arylphosphine oxide moiety on 12 provided appropriately
substituted naphthalene ring 13.7 As in our previously re-
ported synthesis, the presence of the hydroxyl group on
the naphthalene ring is crucial for the optical resolution of
the forthcoming biaryl axis. Out of a screen of several
auxiliaries, naproxen-derived alcohol 148 was found to be
Within the realm of transition-metal-mediated asymmet-
ric processes, the utility of chiral biaryl diphosphine
ligands has been widely demonstrated.1 2,2¢-Bis(diphe-
nylphosphino)-1,1¢-binaphthyl (BINAP; 1) is one of the
most commonly utilized chiral biaryl phosphines avail-
able, and its modification has been well documented.2 In
2005, our group reported the first modification of the 3-
and 3¢-positions of BINAP.3
It was postulated that modifications made at the 3 and 3¢
positions of BINAP would most drastically affect the ster-
ic and electronic properties of the ligand due to the prox-
imity of these positions to the phosphorus atoms. To this
end, we demonstrated that ligands 2–5 outperformed
BINAP in the Rh-catalyzed hydrogenation of N-acetami-
do acrylic acid and its methyl ester, providing enantiose-
lectivities of up to >99% ee for the resulting alanines
(Figure 1). Also included in our initial report was the hy-
drogenation of methyl N-acetamido cinnamate using 1–5.
Performing this hydrogenation with BINAP led to poor
enantioselectivity of the corresponding phenyl alanine but
SYNLETT 2009, No. 15, pp 2513–2517
x
x
.x
x
.2
0
0
9
Advanced online publication: 27.08.2009
DOI: 10.1055/s-0029-1217741; Art ID: S04809ST
© Georg Thieme Verlag Stuttgart · New York

2514
D. A. Rankic et al.
LETTER
OH
OH
OP(O)Ar2
ClP(O)Ar2
LDA
DMAP, Et3N
CH2Cl2
–78 °C to r.t.
18 h, 85%
85%
Br
Br
11
12
HO
O
Np-OMe
14
DEAD, Ph3P
OMe
P(O)Ar2
P(O)Ar2
15
toluene, r.t., 18 h, 75%
Br
Br
13
1. Cu powder
DMF, 140 °C
3 h, 74%
Np-OMe =
Ar =
2. optical resolution
OMe
O
Np-OMe
HSiCl3, Bu3N
P(O)Ar2
P(O)Ar2
(Rax)-7 (62%)
(Sax)-8 (74%)
xylenes
140 °C, 24 h
Np-OMe
O
(Rax)-16 (29%)
(Sax)-17 (36%)
Scheme 1 Synthesis of diastereomeric 3,3¢-disubstituted xylBINAP ligands (Rax)-7 and (Sax)-8
most suitable for our purposes. Auxiliary 14 was coupled diphosphine oxide 18 and hence the xylBINAP derivative
to naphthol 13 using standard Mitsunobu conditions.9 (R)-9 after phosphine oxide reduction (Scheme 2).11
Heating ether 15 in the presence of copper power facilitat-
Chiral auxiliary cleavage from (Sax)-16 followed by acy-
ed biaryl bond formation to afford two diastereomeric
lation with trimethylacetyl chloride provided the diphos-
diphosphine oxides (Rax)-16 and (Sax)-17, which could be
phine oxide of (S)-10, but reduction of this material
resolved using column chromatography10 followed by re-
resulted in a complex mixture of compounds with varying
states of phosphine oxidation and ester cleavage. To cir-
crystallization. Diphosphine oxides (Rax)-16 and (Sax)-17
were reduced independently with HSiCl3 to afford diaste-
cumvent this problem, after chiral auxiliary cleavage, di-
reomeric 3,3¢-disubstituted xylBINAP derivatives (Rax)-7
naphthol (S)-19 was isolated and then reduced under
and (Sax)-8. The absolute stereochemistry of the axis of
chirality for (Sax)-8 was obtained by X-ray crystallogra-
phy of the corresponding Rh(I)(nbd)BF4 adduct; the con-
figuration assignment obtained for (Sax)-8 was extended to
the remainder of ligands in this report.
standard conditions.12 Acylation of the crude reduction
product afforded (S)-10 (Scheme 3).
In order to assess the efficacy our new 3,3¢-disubstituted
xylBINAP derivatives 7–10 as ligands in Rh-catalyzed
asymmetric hydrogenations, it was imperative that any
potential 3,5-dialkyl meta effect was separated from the
effect exerted by increased steric bulk in the 3 and 3¢ po-
sitions. Thus, BINAP (1) and (R)-xylBINAP (6) were
used as controls in order to establish the base 3,5-dialkyl
meta effect.
Synthetic routes to ligands (R)-9 and (S)-10 required re-
moval of the chiral auxiliary, which was facilitated using
BBr3. Cleavage of the chiral auxiliary from (Rax)-16 fol-
lowed by alkylation with 2-bromopropane afforded
O
O
HSiCl3
NBu3
P(O)Ar2
P(O)Ar2
1. BBr3
(Rax)-16
PAr2
PAr2
xylenes
140 °C
24 h, 73%
2. 2-bromopropane
DMF, K2CO3
64% (two steps)
O
O
(R)-18
(R)-9
Scheme 2 Synthesis of (Rax)-3,3¢-(Oi-Pr)2-xylBINAP (9)
Synlett 2009, No. 15, 2513–2517 © Thieme Stuttgart · New York

LETTER
3,3¢-Disubstituted xylBINAP Derivatives
2515
OH
O
1. HSiCl3, NBu3
xylenes
O
BBr3
85%
140 °C, 24 h
P(O)Ar2
P(O)Ar2
PAr2
(Sax)-17
2. ClC(O)t-Bu, Et3N
DMAP, CH2Cl2
24 h, 19%
PAr2
O
(two steps)
OH
O
(S)-19
(S)-10
Scheme 3 Synthesis of (Sax)-3,3¢-(OCOt-Bu)2-xylBINAP (10)
Our initial examination began with the Rh-catalyzed hy- BINAP relative (S)-4. These results reinforce the idea that
drogenation of cyclic enamide 20 using BINAP deriva- changing the steric environment provided by the chiral
tives 2–5 and xylBINAP derivatives 7–10 (Table 1). ligand can lead to improvements in enantioselectivity.
Control hydrogenations using (S)-BINAP and (R)-xyl- However, this does not provide strong evidence that the
BINAP provided amide 21 in low ee’s (16.8% and 13.2%, 3,5-dialkyl meta effect and 3,3¢-disubstitution are operat-
respectively). Of the BINAP derivatives screened, all but ing in a synergistic fashion, but rather that one may be
ligand 2 showed marked improvements over the parent having a stronger influence while the other lies dormant
BINAP, with ligand 5 [OC(O)t-Bu groups in the 3 and 3¢ [i.e. 3,5-dialkyl meta effect dominates in (R)-9 but is di-
positions] showing the most promising result, providing minished when using (R)-10].
21 in 57.4% ee. Employing xylBINAP derivatives 7–10
The introduction of chiral entity 14 into the 3 and 3¢ posi-
provided no further enhancements to the observed ee for
tions of xylBINAP brings forth the possibility of double
compound 21. The best result out of this series was 48.7%
diastereodifferentiation in the enantio-determining step of
ee, provided by (R)-10, which was nearly 10% lower in
these asymmetric hydrogenations. Ligands (Rax)-7 and
selectivity than that achieved by its BINAP analogue 5.
(Sax)-8 did indeed show this phenomenon. Thus, simply
It is clear from the hydrogenation of enamide 20 that the changing the configuration of the chiral axis from R to S,
xylyl groups in synthesized ligand 10 gave no additional without disturbing the configuration of the attached auxil-
improvement in enantioselectivity when compared to iaries, improves the enantioselectivity of the hydrogena-
ligand 5. Failure of 10 to further improve levels of enanti-
oselectivity achieved by 3,3-disubstituted BINAP deriva-
Table 1 Asymmetric Hydrogenation of Cyclic Enamide 20 Using
tive 5 can be attributed to a slight negative 3,5-dialkyl
meta effect,13 which was observed in the comparison of
the control ligands (1 vs 6, vide supra). It has been previ-
ously reported that too much steric bulk around the metal
center may result in partial dissociation of either the
diphosphine ligand or the substrate, resulting in dimin-
ished chirality transfer between the metal catalyst and
substrate.14
BINAP Ligands 1–5 and xylBINAP Ligands 6–10
NHAc
NHAc
Rh(nbd)2BF4 (1 mol%)
ligand (1.1 mol%), MeOH
H2 (2 atm), r.t., 24 h
21
20
Entrya,b
Ligand
(S)-1
eec (%)
16.8 (S)
5.2 (R)
20.3 (R)
35.3 (S)
57.4 (R)
13.2 (R)
33.9 (R)
20.9 (S)
36.9 (S)
48.7 (S)
Next, we moved on to the Rh-catalyzed hydrogenation of
methyl N-acetamido cinnamate (22). When xylBINAP 6
was used as a ligand in the hydrogenation of 22, it provid-
ed an enantioselectivity of 66.5%, nearly five times great-
er than that provided by BINAP (1; Table 2, entries 1 and
6), demonstrating that the 3,5-dialkyl meta effect is oper-
ational in this system. Our previous results showed that
(S)-2 and (S)-4 were the best ligands out of the BINAP se-
ries and provided enantioselectivities of 23.2% and
74.7%, respectively, for the hydrogenation of 22.3
1
2
(S)-2
3
(S)-3
4
(S)-4
5
(S)-5
6
(R)-6
(Rax)-7
(Sax)-8
(R)-9
(R)-10
7
In order to obtain a direct comparison to BINAP deriva-
tives 2 and 4, we chose to examine the xylBINAP ana-
logues of 2 and 4 in the Rh-catalyzed hydrogenation of 20.
Using (R)-9 as a ligand, provided 23 in 69.6% ee, repre-
senting a three-fold increase in selectivity when compared
to the related BINAP ligand (S)-2, but comparable to the
8
9
10d
a All reactions proceeded with complete conversion.
result obtained with (R)-6 (entries 2, 4 and 7; Table 2). b Catalyst formed in situ by premixing ligand with Rh(nbd)2BF4 in
Ligand (R)-10 provided phenyl alanine 23 in 73.8% ee,
which was slightly higher than that provided by control 6
but comparable to that obtained with 3,3-disubstituted
methanol for 30 min.
c Determined by chiral GC using a Cyclodex B column.
d The opposite configuration of ligand to that shown in Scheme 3 was
used.
Synlett 2009, No. 15, 2513–2517 © Thieme Stuttgart · New York

2516
D. A. Rankic et al.
LETTER
27% conversion after seven hours (Table 3, entry 4). A
similar temperature study using 8 revealed selectivities as
high as >99% ee at –20 °C with 53% conversion (see en-
try 7). A synthetically useful 90% conversion and 95.8%
ee was obtained at 0 °C (entry 6).
Table 2 Asymmetric Hydrogenation of Methyl N-Acetamido Cin-
namic Acid (22) Using BINAP Ligands 1–5 and xylBINAP Ligands
6–10
Rh(nbd)2BF4 (1 mol%)
CO2Me
NHAc
CO2Me
NHAc
ligand (1.1 mol%), MeOH
H2 (2 atm), r.t., 7 h
Ph
Ph
Lower temperatures for these hydrogenations not only
provide higher enantioselectivities, but they also provide
us with mechanistic insights. Traditionally, this type of
Rh-catalyzed hydrogenation undergoes an ‘anti-lock and
key’ mechanism in which it is the minor Rh-olefin diaste-
reomer that undergoes hydrogenation.15 This type of
mechanism leads to higher enantioselectivities with high-
er temperatures. Our temperature studies reinforce the
idea that these novel 3,3′-disubstituted BINAP and xylBI-
NAP derivatives cause the Rh-catalyzed hydrogenation of
22 to undergo a mechanism different from that which was
previously established.16
23
22
Entrya,b
Ligand
(S)-1
%eec
1d
2d
3d
4
14.8 (R)
23.2 (S)
74.7 (S)
66.5 (R)
80.1 (R)
91.0 (S)
69.6 (R)
73.8 (R)
(S)-2
(S)-4
(R)-6
5
(Rax)-7
(Sax)-8
(R)-9
6
In summary, we have developed a synthetic route to 3,3′-
disubstituted xylBINAP ligands. We have demonstrated
that the additional steric bulk present in these ligands can
either improve or shut down the selectivity of a given hy-
drogenation, depending on the substitution pattern of the
ligand and the hydrogenation substrate. The introduction
of chiral entities to the 3 and 3′ positions have also proven
to be highly effective for the hydrogenation reactions.17
We have provided more evidence for a different mecha-
nistic mode for Rh-catalyzed hydrogenations when this
class of ligands is utilized. We are exploring new ways of
substituting the BINAP skeleton and are continuing to
probe the mechanism of Rh-catalyzed hydrogenations us-
ing this novel class of diphosphine ligands.
7
8e
(R)-10
a All reactions proceeded with complete conversion.
b Catalyst formed in situ by premixing ligand with Rh(nbd)2BF4 in
methanol for 30 min.
c Determined by chiral HPLC using a Chiralcel OD column.
d Entries reproduced from ref 3 for purposes of comparison.
e The opposite configuration of ligand to that shown in Scheme 3 was
used.
Table 3 Temperature Studies on the Hydrogenation of 22 Using
BINAP Ligand 4 and xylBINAP Ligand 8
Entrya
Ligand
(S)-4
Temp (°C) Conv (%)b %eec
1
2
3
4
5
6
7
8
25
0
100
84
82
27
100
90
53
0
71.4 (S)
73.1 (S)
69.2 (S)
80.5 (S)
91.0 (S)
95.8 (S)
>99 (S)
–
Supporting Information for this article is available online at
http://www.thieme-connect.com/ejournals/toc/synlett.
(S)-4
(S)-4
–20
–45
25
Acknowledgment
(S)-4
Financial support for this work was provided by Merck Frosst
Canada and by the Natural Sciences and Engineering Research
Council of Canada through a CRD grant to B.A.K., and NSERC
postgraduate scholarships to both D.A.R. and J.M.H. Alberta
Ingenuity is also thanked for studentships to D.A.R. and J.M.H.
Thanks are also extended to Prof. Warren Piers for use of his chiral
HPLC.
(Sax)-8
(Sax)-8
(Sax)-8
(Sax)-8
0
–20
–45
a Catalyst formed in situ by premixing ligand with Rh(nbd)2BF4.
b Determined by GC–MS.
References and Notes
c Determined by chiral HPLC on a Chiralcel OD column.
(1) (a) McCarthy, M.; Guiry, P. J. Tetrahedron 2001, 57, 3809.
(b) Tang, W.; Zhang, X. Chem. Rev. 2003, 103, 3029.
(c) Shimizu, H.; Nagasaki, I.; Saito, T. Tetrahedron 2005,
61, 5405. (d) Li, Y.-M.; Kwong, F.-Y.; Yu, W.-Y.; Chan,
A. S. C. Coord. Chem. Rev. 2007, 251, 2119.
tion of 22 from 80.1% to 91.0% (Table 2, entry 5 vs.
Table 3, entry 5). Furthermore, both (Rax)-7 and (Sax)-8
outperform related ligands 2–5, 9, and 10, and provide the
highest ee’s reported for this transformation with this
class of BINAP ligands. This is interesting in itself, as the
chiral centers of the pendant auxiliary are far removed
from the phosphorus atoms in the solid state when these
ligands are bound to RhI. We also found that lowering the
temperature of the hydrogenation to –45 °C using 4, af-
forded phenyl alanine 23 in 80.5% ee, albeit with only
(2) Berthod, M.; Mignani, G.; Woodward, G.; Lemaire, M.
Chem. Rev. 2005, 105, 1801.
(3) Hopkins, J. M.; Dalrymple, S. A.; Parvez, M.; Keay, B. A.
Org. Lett. 2005, 7, 3765.
(4) For seminal publications, see: (a) Trabesinger, G.; Albinati,
A.; Feiken, N.; Kunz, R.; Pregosin, P. S.; Tschoerner, M.
J. Am. Chem. Soc. 1997, 119, 6315. (b) Selvakumar, K.;
Valentini, M.; Pregosin, P. S.; Albinati, A.; Eisenträger, F.
Organometallics 2000, 19, 1299.
Synlett 2009, No. 15, 2513–2517 © Thieme Stuttgart · New York

LETTER
3,3¢-Disubstituted xylBINAP Derivatives
2517
(5) Newman, M. S.; Sankaran, V.; Olson, D. R. J. Am. Chem.
Soc. 1976, 98, 3237.
(14) For a detailed study on a bulky ligand, see: Imamoto, I.;
Hoge, G.; Kouchi, M.; Takahashi, H. J. Am. Chem. Soc.
2008, 130, 2560.
(6) Mashima, K.; Kusano, K.; Sato, N.; Matsumara, Y.; Nozaki,
K.; Kumobayashi, H.; Sayo, N.; Hori, Y.; Ishizaki, T.;
Akutagawa, S.; Takaya, H. J. Org. Chem. 1994, 59, 3064.
(7) This type of regioselective migration has been observed
previously for the non-brominated naphthalene, see:
Dhawan, B.; Redmore, D. J. J. Org. Chem. 1991, 56, 833.
(8) This alcohol is available from standard LiAlH4 reduction
of (+)-(S)-2-(6-methoxynaphthalen-2-yl)propanoic acid
[(S)-naproxen].
(9) Attempts at forming the ether linkage via alkylative methods
failed and resulted in elimination rather than substitution
(10) (Sax)-8, Rf = 0.35; (Rax)-7, Rf = 0.25 (EtOAc–hexanes, 35%)
(11) Cleavage of the chiral auxiliary followed by alkylation with
MeI successfully afforded 3,3¢-(MeO)2-xylBINAP(O).
However, exposing 3,3¢-(MeO)2-xylBINAP(O) to the
reduction conditions caused demethylation. Use of
alternative reduction conditions (AlH3·THF) resulted in
decomposition.
(12) We presumed that the sensitivity of the ester moieties was
due to the increase of steric bulk around the phosphorus
atoms. Reducing phosphine oxides in the presence of free
hydroxy groups is known in the literature for sterically
encumbered phosphorus atoms, see: Paruch, K.; Vyklický,
L.; Wang, D. Z.; Katz, T. J.; Incarvito, C.; Zakharov, L.;
Rheingold, A. L. J. Org. Chem. 2003, 68, 8539.
(13) For selected examples of other systems that have exhibited a
negative 3,5-dialkyl meta effect, see: (a) Hatano, M.;
Terada, M.; Mikami, K. Angew. Chem. Int. Ed. 2001, 40,
249. (b) Dotta, P.; Kumar, P. G. A.; Pregosin, P. S. Magn.
Reson. Chem. 2002, 40, 653.
(15) Landis, C. R.; Halpern, J. J. Am. Chem. Soc. 1987, 109,
1746.
(16) For a review on the mechanism of enantioselection in Rh-
catalyzed hydrogenations, see: Gridnev, I. D.; Imamoto, T.
Acc. Chem. Res. 2004, 37, 633.
(17) General Procedure for Rh-Catalyzed Asymmetric
Hydrogenations: Rh(nbd)2BF4 (3.7 mg, 0.01 mmol), ligand
(0.011 mmol) and MeOH (1 mL) were combined under an
inert atmosphere and stirred for 30 min at room temperature.
Olefin (1 mmol) was added with an additional aliquot of
MeOH (1 mL). The reaction mixture was then subjected to
three consecutive freeze/pump/thaw cycles on a double
manifold (pump time = 10 min). Upon warming to room
temperature, the reaction was placed under a hydrogen
atmosphere (2 atm) for 7 h. Upon completion, volatiles were
removed in vacuo and the remaining residue was passed
through a plug of silica gel (hexanes–EtOAc, 1:1). Upon
solvent evaporation, the isolated product was used directly
for either chiral GC or chiral HPLC analysis.
Chiral GC data for N-(1,2,3,4-tetrahedronaphthalen-1-
yl)acetamide(21): Cyclodex B column; isothermal 170 °C;
tR1 [(R)-21] = 23.5 min, tR2 [(S)-21] = 24.4 min.
Chiral HPLC data for methyl a-acetamido cinnamate (23):
Chiral HPLC, Chiralcel OD; hexane–i-PrOH, 9:1; 1.0 mL/
min; tR1 [(R)-23] = 12.4 min, tR2 [(S)-23] = 16.3 min.
Synlett 2009, No. 15, 2513–2517 © Thieme Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.