1H and13C nuclear magnetic resonance studies of the hindered phencyclone adducts of some smaller branched N-alkyl maleimides: Rigorous aryl proton assignments with high-resolution two-dimensional (COSY45) spectroscopy, and anisotropic shielding effects and ab initio geometry optimizations

Article abstract of DOI:10.1366/0003702053585426

Phencyclone, 1, a potent Diels-Alder diene, reacts with a series of N-alkylmaleiniides, 2, to form hindered adducts, 3. The 300 MHz ¹H and 75 MHz ¹³C NMR studies of these adducts at ambient temperatures have demonstrated slow rotations on the nuclear magnetic resonance (NMR) timescales for the unsubstituted bridgehead phenyl groups, and have revealed substantial magnetic anisotropic shielding effects in the ¹H spectra of the N-alkyl groups of the adducts. The selected N-alkyl groups for the target compounds emphasized smaller branched alkyls, including C₃ (isopropyl, a); C₄ (isobutyl, b; and t-butyl, c); C₅ (n-pentyl, d; isopentyl [isoamyl], e; 1-ethylpropyl, f; t-amyl, g;) and a related C₈ isomer (1,1,3,3-tetramethylbutyl ["t-octyl"], h). The straight-chain n-pentyl analog was included as a reference. This present work on the branched N-al-kylmaleimide adducts appreciably extends our earlier compilation on the N-n-alkylmaleimide adducts. Key methods for proton assignments included "high-resolution" ¹H-¹H chemical shift correlation spectroscopy, COSY45. ¹³C NMR of the adducts, 3, verified the expected number of aryl carbons for slow exchange limit (SEL) spectra of the bridgehead phenyl groups. The synthetic routes involved reaction of the corresponding amines, 4, with maleic anhydride to give the N-alkylmaleamic acids, 5, which underwent cyclodehydration to form the maleimides, 2. Magnetic anisotropic shielding magnitudes for alkyl group protons in the adducts were calculated relative to corresponding proton chemical shifts in the maleimides. Geometry optimizations for the above adducts (and for the N-n-butylmaleimide adduct) were performed at the Hartree-Fock level with the 6-31G* basis set. The existence of different contributing conformers for the adducts is discussed with respect to their calculated energies and implications regarding experimentally observed anisotropic shielding magnitudes.

Full text of DOI:10.1366/0003702053585426

13
1H and C Nuclear Magnetic Resonance Studies of the
Hindered Phencyclone Adducts of Some Smaller Branched
N-Alkyl Maleimides: Rigorous Aryl Proton Assignments
with High-Resolution Two-Dimensional (COSY45)
Spectroscopy, and Anisotropic Shielding Effects and Ab
Initio Geometry Optimizations
RONALD CALLAHAN, RON PRIP, NAVROZ SHARIFF, OLGA SKLYUT, ROBERT
ROTHCHILD,* and KEVIN BYNUM†
New York University, Chemistry Department, 569 Brown, 29 Washington Place, New York, New York 10003 (R.C.); The City
University of New York, John Jay College of Criminal Justice, Science Department, 445 West 59th Street, New York, New York
10019-1128 (R.P., N.S., O.S., R.R., K.B.); and Doctoral Faculty, Graduate School and University Center, City University of New
York, New York, New York (R.R.)
Phencyclone, 1, a potent Diels-Alder diene, reacts with a series of
N-alkylmaleimides, 2, to form hindered adducts, 3. The 300 MHz
1H and 75 MHz 13C NMR studies of these adducts at ambient tem-
1, with N-alkylmaleimides, 2, have previously discussed
the products from straight-chain, N-n-alkylmaleimides,
including the adducts from N-n-propylmaleimide,1 N-n-
peratures have demonstrated slow rotations on the nuclear mag-
butylmaleimide,2 N-n-hexylmaleimide,3 and the longer
netic resonance (NMR) timescales for the unsubstituted bridgehead
chains up through N-n-dodecylmaleimide.4 This present
phenyl groups, and have revealed substantial magnetic anisotropic
work on the branched N-alkylmaleimide adducts appre-
1
shielding effects in the H spectra of the N-alkyl groups of the ad-
ciably extends our earlier compilation on the N-n-alkyl-
maleimide adducts.4 Phencyclone is a potent Diels-Alder
diene5 that readily forms adducts, 3, from electrophilic
dienophiles, including maleimides. The maleimide ad-
ducts of phencyclone are notably hindered, such that the
unsubstituted bridgehead phenyl groups in the adducts
exhibit slow exchange limit (SEL) proton or carbon-13
NMR spectra at ambient temperatures with a spectrom-
eter of medium field strength, e.g., 7.05 tesla (300 MHz
1H and 75 MHz 13C). Thus, the rotation of the phenyls
about the C(sp3)–C(aryl sp2) bonds is slow on the NMR
timescales. [Throughout this paper, we refer to the ‘‘NMR
timescales,’’ which are actually defined by (i.e., related
to the reciprocal of) the difference in chemical shifts, ⌬␯
(expressed in Hz), between the corresponding nuclei of
the interconverting conformers. These ⌬␯ magnitudes
will be proportional to the NMR spectrometer field
strength (which defines the observation frequencies of the
nuclei), and will depend upon the actual nuclei and spe-
cific sample conditions, e.g., solvent, temperature, etc.]
The severe hindrance appears to result most strongly
ducts. The selected N-alkyl groups for the target compounds em-
phasized smaller branched alkyls, including C3 (isopropyl, a); C4
(isobutyl, b; and t-butyl, c); C5 (n-pentyl, d; isopentyl [isoamyl], e;
1-ethylpropyl, f; t-amyl, g;) and a related C8 isomer (1,1,3,3-tetra-
methylbutyl [‘‘t-octyl’’], h). The straight-chain n-pentyl analog was
included as a reference. This present work on the branched N-al-
kylmaleimide adducts appreciably extends our earlier compilation
on the N-n-alkylmaleimide adducts. Key methods for proton assign-
1
ments included ‘‘high-resolution’’ H–1H chemical shift correlation
spectroscopy, COSY45. 13C NMR of the adducts, 3, verified the ex-
pected number of aryl carbons for slow exchange limit (SEL) spec-
tra of the bridgehead phenyl groups. The synthetic routes involved
reaction of the corresponding amines, 4, with maleic anhydride to
give the N-alkylmaleamic acids, 5, which underwent cyclodehydra-
tion to form the maleimides, 2. Magnetic anisotropic shielding mag-
nitudes for alkyl group protons in the adducts were calculated rel-
ative to corresponding proton chemical shifts in the maleimides.
Geometry optimizations for the above adducts (and for the N-n-
butylmaleimide adduct) were performed at the Hartree-Fock level
with the 6-31G* basis set. The existence of different contributing
conformers for the adducts is discussed with respect to their cal-
culated energies and implications regarding experimentally ob-
served anisotropic shielding magnitudes.
Index Headings: Hindered rotations; Magnetic anisotropy; Nuclear
magnetic resonance; Dynamic NMR; One- and two-dimensional (1D
and 2D) NMR; COSY; Hartree-Fock; Molecular modeling.
from potential interactions of the ortho protons, H2Ј,6Ј,
of the bridgehead phenyls with the H1,8 hydrogens of
the phenanthrenoid moiety in the adducts. Compound
structures and atom numbering are shown in Fig. 1. There
are additional contributing interactions between the ortho
protons and: (a) the carbonyl oxygens in the imide moiety
of the adducts; (b) the bridging ketone carbonyl; and (c)
the bridgehead methines of the adducts. In addition to
allowing the examination of this hindered rotation, NMR
studies of the adducts of 1 with 2 have revealed rather
dramatic magnetic anisotropic effects, particularly with
respect to shielding of some protons in the N-alkyl group
of the adducts.1,2 This latter shielding effect can be attri-
buted to the position of the alkyl groups in the expected
INTRODUCTION
Our ongoing nuclear magnetic resonance (NMR) stud-
ies of the hindered Diels-Alder adducts of phencyclone,
Received 11 May 2004; accepted 8 November 2004.
* Author to whom correspondence should be sent. E-mail: rrothchild@
jjay.cuny.edu.
† Current address: Eyetech Pharmaceuticals, 140 East Hanover Avenue,
Cedar Knolls, NJ 07927.
0003-7028/05/5903-0354$2.00 /0
2005 Society for Applied Spectroscopy
354
Volume 59, Number 3, 2005
APPLIED SPECTROSCOPY
᭧

EXPERIMENTAL
General NMR and other techniques were described
earlier.1–3 Spectra were obtained on a Bruker ACF300
1
NMR spectrometer (7.05 tesla) at 300 MHz for H and
75 MHz for 13C, equipped with QNP (quad nuclear
probe) and Aspect 3000 data system, at ambient temper-
atures in CDCl3 unless otherwise noted. Proton shifts
were referenced to internal tetramethylsilane (TMS) at
0.0 ppm. For carbon-13 spectra, shifts were referenced to
the center line of the CDCl3 triplet at 77.0 ppm; in d6-
acetone, shifts were referenced to the center line of the
d6-acetone septet at 29.8 ppm. Where proton decoupling
was applied for carbon-13 spectra, WALTZ16 was used
for composite pulse decoupling. Standard Bruker micro-
programs were routinely employed for two-dimensional
(2D) chemical shift correlation spectra. Typical parame-
ters for the ‘‘high-resolution’’ COSY45 spectra of the ad-
duct aryl regions used two dummy scans and 16 tran-
sients for each of 256 t1 increments with a spectral width
from ca. 7.0 to 8.8 ppm. Reagents and solvents were
obtained from Aldrich Chemical (Milwaukee, WI) or
Lancaster Synthesis (Windham, NH) and were used with-
out further purification. Infrared spectra were acquired
with a Perkin Elmer model 1640 instrument with DTGS
1
(deuterated triglycine sulfate) detector and 4 cmϪ reso-
lution, using KBr pellets or disposable polyethylene cards
(3M, type 61); in the latter case, peaks in the 2800–3000
1
cmϪ region have been omitted from the listed peaks be-
FIG. 1. Summary of synthetic routes with general compound structures
cause of matrix absorptions. Only major peaks have been
reported. Reported melting points are uncorrected. Ge-
ometry optimization calculations were performed using
Spartan ’04 ESSENTIAL (v. 2.0.0, Wavefunction, Inc.
and atom numbering.
endo isomer from the Diels-Alder addition. The alkyl
group can be directed ‘‘into’’ the cavity of the phency-
clone adduct, thereby being directed ‘‘towards’’ the phen-
anthrenoid moiety, from which the anisotropic shielding
could arise. The magnetic anisotropic shielding effects
referred to here characterize the influence of functional
groups and structural features in the sample molecules
that result in notable shielding (or deshielding) effects at
specific nuclei in the molecule as it tumbles in the solu-
tions in which the NMR spectra are acquired. (This tum-
bling in liquid solutions results in observation of the iso-
tropic averages of shielding tensors (or chemical shift
tensors) that are anisotropic.)
We have recently begun to apply computational tech-
niques, performing ab initio geometry optimizations at
the Hartree-Fock level, using the 6-31G* basis set, to get
some insight into adduct structures.4 For some of the N-
n-alkyl systems that we have investigated, a qualitative
explanation of the magnetic anisotropic shielding mag-
nitudes considered different possible conformers of the
n-alkyl group in the adducts. To extend our understanding
of these systems, we report here our investigations using
systems in which the N-alkyl groups are variously
branched rather than straight-chain. The alkyl groups ex-
amined in this present report include: C3 (isopropyl, a);
C4 (isobutyl, b; and t-butyl, c); C5 (n-pentyl, d; isopentyl
[isoamyl], e; 1-ethylpropyl, f; t-amyl, g;); and a related
C8 isomer (1,1,3,3-tetramethylbutyl [‘‘t-octyl’’], h). The
straight-chain n-pentyl analog was included as a compar-
ative reference. The N-n-butylmaleimide adduct was in-
cluded for computational comparison.
᭧1991–2003; October 8, 2003) on a Dell Pentium 4 plat-
form with 2.26 GHz processor speed and 512 MB mem-
ory. Calculation parameters included the disabling of
symmetry, and turning convergence on.
Preparation of N-Alkylmaleamic Acids. General
method:3 A solution of the required amine in CH2Cl2 (ca.
5 mL/g amine) was added dropwise, over 10 min, to a
stirred chilled slurry of maleic anhydride (equimolar with
amine) in CH2Cl2 (ca. 3 mL/g anhydride) at 0–5Њ. The
initially undissolved maleic anhydride was consumed and
went into solution during or shortly after addition of the
amine. For the less hindered amines, the mixture was
allowed to warm to room temperature and stirred for 30
min. If the crude maleamic acid spontaneously crystal-
lized, it could be collected by vacuum filtration, washed
with cold CH2Cl2, and dried for use in conversion to the
maleimide. In some cases, crude maleamic acid product
did not spontaneously crystallize, and solvent was re-
moved (rotary evaporator, aspirator pressure, ca. 50Њ bath
temperature) to give a pale yellow oil. On standing (or
after trituration), the oil crystallized and could be washed
with cold CH2Cl2, collected by vacuum filtration and
washed with additional cold solvent, and dried. Typical
yields were 60–80%. For the more hindered amines, such
as t-amylamine, the reaction mixture of amine and maleic
anhydride was refluxed in CH2Cl2 for 2 h prior to workup
as above. For the (1,1,3,3-tetramethyl)butylamine (‘‘t-oc-
tyl’’ amine), a solution of the amine in toluene was added
dropwise with stirring to (equimolar) maleic anhydride in
toluene, maintained during addition at 45–50
Њ, and the
APPLIED SPECTROSCOPY
355

mixture was maintained at this temperature for approxi-
mately 2 h, then cooled to room temperature. The crude
maleamic acid was collected by vacuum filtration,
washed with cold solvent, and dried.
oil.8 (1H NMR): 6.63 (2H, s, HC
6.95, NCH); 1.39 (6H, d, J
of the acetoxypyrrolidinedione adduct impurity was sug-
ϭ
CH); 4.34 (1H, septet,
J
ϭ
ϭ 6.93, Me2). About 30%
3
gested by peaks at: 5.37 (1H, dd, J
ϭ
8.82 and 4.78,
8.82, HM); 2.63 (1H,
ϭ 18.2, J ϭ 4.96, HA); 2.17 or 2.10 [tentative]
Preparation of N-Alkylmaleimides. General method:3
The crude maleamic acid, prepared as above, typically
ca. 10–30 mmol, was combined with 10–30 mL acetic
anhydride and ca. 1–3 g anhydrous sodium acetate in a
flask equipped with reflux condenser topped with a dry-
ing tube. (About 1 mmol maleamic acid to 1 mL Ac2O
and 0.1 g NaOAc was a usual ratio.) After heating on a
boiling water bath for 0.5–2 h, during which time the
mixture turned a dark red-black, the mixture was cooled,
poured into ice slush, and extracted with CH2Cl2. The
organic layer was extensively washed with water and
aqueous NaHCO3 and dried, and then the solvent was
removed by rotary evaporator. The crude maleimides were
obtained as oils, estimated 20–35% yields by proton NMR,
but were suitable for conversion to the phencyclone adducts
without further purification. For the ‘‘t-octyl’’mal-
eimide, i.e., N-(1,1,3,3-tetramethyl)butylmaleimide, the
maleamic acid was combined with 2 equivs triethylamine
in acetone, and maintained at reflux, during which time
1.5 equivs acetic anhydride was added dropwise. After
overnight reflux, cooling, and quenching with ice water,
the mixture was stirred for several hours. The resulting
crude dark tarry product was separated and partitioned
between aqueous ethanol and hexanes. The hexane layer
was separated from the dark aqueous alcohol layer; grad-
ual evaporation of the hexanes resulted in crystals of the
maleimide, ca. 15–25% yield.
HX ); 3.11 (1H, dd, J ϭ ϭ
18.38, 3J
2
2
3
dd, J
(3H, s, AcO). The possible presence of some (ca. 4%)
(trans) fumaramic acid impurity may have been indicat-
ed: 6.97 (1H, d, J
15.44, half of AB q) suggesting the trans HC
moiety.
N-(Isopropyl)maleimide Adduct with Phencyclone.
ϭ
15.45, half of AB q); 6.81 (1H, d,
J
ϭ
ϭCH
1
Mpr 264–267
Њ
(dec.) IR (polyethylene card, cmϪ ):
1789.0 (strained ketone); 1700.1 (NC
ϭ
O); 1653.0;
1545.0; 1499.1; 1447.3; 1362.0; 1227.3; 1128.0; 779.4;
753.0; 724.0; 696.3. (13C NMR, CDCl3): 197.04 (ketone);
174.33 (NCϭO); 133.88 (Q); 133.35 (Q); 131.18 (Q);
130.91 (2CH); 129.32 (2CH); 129.09 (2CH); 128.60
(2CH); 128.36 (2CH); 127.00 (2CH); 126.56 (2CH);
126.37 (Q); 125.90 (2CH); 122.90 (2CH); 63.36 (C6H5C);
43.98 (bridgeheads); 43.72 (NCH); 17.75 Me2).
Isobutylamine. (1H NMR:) 2.49 (2H, d, J
ϭ 6.60,
NCH2); 1.57 (1H, nonet, J
br s, NH2); 0.90 (6H, d, J
ϭ
ϭ
6.63, NCH2CH); 1.11 (2H,
6.64, Me).
N-(i-Butyl)maleamic Acid. Mpr 88–93
100Њ.9 (1H NMR:) 7.99 (1H, v br t, NH); 6.53 (1H, d, J
12.78, half of AB q, HC CH); 6.32 (1H, d, J 12.79,
half of AB q, HC CH); 3.21 (2H, dd [nearly t], J 6.78
and 6.27, NCH2); 1.91 (1H, m, J ca. 6.76, NCH2CH);
0.96 (6H, d, J 6.70, Me2).
N-(i-Butyl)maleimide. Oil, lit. mp 47–48Њ.9 (1H NMR:)
6.70 (2H, s, HC CH); 3.34 (2H, d, J 7.35, NCH2);
2.02 (1H, m, J 6.91, NCH2CH); 0.89 (6H, d, J 6.99,
Me2). The presence of the impurity acetoxypyrrolidine-
Њ, lit. mp 99–
ϭ
ϭ
ϭ
ϭ
ϭ
ϭ
Preparation of Phencyclone Adducts of N-Alkyl-
maleimides. General method:1–3 Ideally, the maleimide
would be used in ca. 5 mol % excess relative to the phen-
cyclone so that discharge of the intense green-black phen-
cyclone color would indicate reaction completion. Where
crude maleimides were used, a nominal two-fold excess
was routinely used to assure consumption of the phen-
cyclone. (A further aliquot of crude maleimide could be
added if required for discharge of phencyclone color.)
The addition reactions were typically run with 1–2 mmol
phencyclone combined with the maleimide in a screw-
ϭ
ϭ
ϭ
ϭ
3
dione (ca. 20%) was indicated by: 5.43 (1H, dd, J
ϭ
2
3
8.82 and 4.78, HX ); 3.17 (1H, dd, J
ϭ
18.38, J
ϭ
8.82,
2
3
HM); 2.70 (1H, dd, J
ϭ 18.2, J ϭ 4.96, HA); 2.17 (3H,
s, AcO). The possible presence of some (ca. 8%) (trans)
fumaramic acid impurity may have been indicated: 6.97
(1H, d, J
half of AB q) suggesting the trans HC
N-(i-Butyl)maleimide Adduct with Phencyclone. Re-
crystallized, mpr 268–275 (dec). IR (polyethylene card,
cmϪ ): 1794.2 strained ketone; 1703.0 (NC
O); 1699.0
(NC O); 1461.9; 1448.2; 1203.8; 1140.0; 872.3; 751.5;
724.5; 701.2. (13C NMR, CDCl3): 196.48 (ketone);
174.68 (NC O); 133.85 (Q); 133.35 (Q); 131.43 (Q);
ϭ
15.45, half of AB q); 6.81 (1H, d, J
ϭ 15.44,
ϭ
CH moiety.
cap vial (Teflon௡-lined cap) of 15–20 mL capacity,
Њ
1
equipped with a magnetic stirbar. CH2Cl2 solvent was
added to within 5 mm of the vial lip (to leave minimal
headspace), and the vial was capped. The less-hindered
maleimides would decolorize the phencyclone after 0.5–
2 h of stirring at room temperatures. For the more hin-
dered maleimides, e.g., t-amyl and ‘‘t-octyl’’, reactions
were run under reflux in a round-bottom flask with reflux
condenser for 6–20 h. For either method, after consump-
tion of phencyclone, solvent was removed and the crude
product was collected by vacuum filtration, washed with
cold CH2Cl2, and dried. Adducts were recrystallized from
pentane–CH2Cl2. Typical isolated yields were 15–30%.
All the adducts melted with decomposition: darkening
and gas evolution were observed.
ϭ
ϭ
ϭ
130.93 (2CH); 129.35 (2CH); 129.16 (2CH); 128.58
(2CH); 128.38 (2CH); 127.02 (2CH); 126.58 (2CH);
126.24 (Q); 125.78 (2CH); 122.93 (2CH); 63.36 (C6H5C);
46.62 (NCH2); 44.57 (bridgeheads); 27.04 (NCH2CH);
19.23 (Me2).
N-(t-Butyl)maleimide. Commercial material (Aldrich)
was used. (1H NMR:) 6.55 (2H, s, HC
s, Me3).10
ϭCH); 1.58 (9H,
N-(t-Butyl)maleimide Adduct with Phencyclone.
1
Mpr 265–267
tone; 1703.3 (NC
NMR, CDCl3): 197.06 (ketone); 175.35 (NC
Њ
(dec). IR (KBr, cmϪ ): 1792.1 strained ke-
O); 1331.5; 1167.5; 752.1; 697.5. (13C
O); 133.97
N-(Isopropyl)maleamic Acid. Mpr 95–102
103Њ.6 (1H NMR): 7.73 (1H, v br d, J
5.88, NH); 6.47
(1H, d, J 12.81, half of AB q, HC CH); 6.31 (1H, d,
12.80, half of AB q, HC CH); 4.14 (1H, approx.
octet, J ca. 6.84, NCH); 1.26 (6H, d, J 6.59, Me2).
N-(Isopropyl)maleimide. Obtained as oil, lit. mp 27Њ 7
Њ
, lit. mp
ϭ
ϭ
ϭ
ϭ
ϭ
(Q); 133.45 (Q); 131.20 (Q); 130.93 (2CH); 129.31
(2CH); 129.06 (2CH); 128.59 (2CH); 128.31 (2CH; note
that any traces of benzene from reaction solvent show up
nearly isochronous with this signal); 126.99 (2CH);
J
ϭ
ϭ
ϭ
,
356
Volume 59, Number 3, 2005

126.55 (2CH overlapped with Q); 125.91 (2CH); 122.95
(2CH); 63.68 (C6H5C); 58.41 (NCMe3); 43.74 (bridge-
head CH); 26.97 (methyls). (13C NMR, CD2Cl2, shifts ref-
erenced to central peak of solvent quintet at 53.8 ppm):
723.5; 697.8. (13C NMR): 196.72 (ketone); 174.28
(NC O); 133.85 (Q); 133.33 (Q); 131.34 (Q); 130.91
ϭ
(2CH); 129.34 (2CH); 129.12 (2CH); 128.59 (2CH);
128.37 (2CH); 126.96 (2CH); 126.58 (2CH); 126.37 (Q);
125.87 (2CH); 122.97 (2CH); 63.37 (C6H5C); 44.41
(bridgehead); 37.21, 34.90 and 24.64 NCH2CH2CH);
21.57 (Me2).
175.55 (NCϭO); 134.60 (Q); 133.92 (Q); 131.68 (Q);
131.34 (2CH); 129.632 (2CH) and 129.620 (2CH), in-
completely resolved; 128.82 (2CH); 128.70 (2CH, essen-
tially isochronous with traces of benzene from reaction
solvent); 127.36 (2CH); 127.00 (Q); 126.69 (2CH);
126.34 (2CH); 123.44 (2CH); 64.10 (C6H5C); 58.54
(NCMe3); 44.11(bridgehead CH); 27.12 (methyls). (13C
NMR, d6-acetone, shifts referenced to central peak of sol-
vent septet at 29.83 ppm): 197.37 (ketone); 176.33
N-(1-Ethylpropyl)maleamic Acid. Mpr 85–92
NMR): 7.58 (1H, br s, NH); 6.53 (1H, br d, J 12.80,
half of AB q, HC CH); 6.31 (1H, d, J 12.74, half of
AB q, HC CH); 3.84 (1H, complex symmet. m, NCH);
1.44–1.70 (4H, very complex mults, diastereotopic pairs
of CH2 protons); 0.93 (6H, t, J 7.44, Me).
N-(1-Ethylpropyl)maleimide. Oil. (1H NMR): 6.65
(2H, s, HC CH); 3.77–3.88 (1H, m, NCH); 1.61–1.80
(4H, m, CH2); 0.82 (6H, t, J 7.32, Me). About 40% of
Њ
. (1H
ϭ
ϭ
ϭ
ϭ
ϭ
(NCϭO); 135.73 (Q); 134.76 (Q); 132.17 (Q); 132.10
(2CH); 130.36 (2CH); 129.76 (2CH); 129.05 (2CH);
128.88 (2CH); 127.77 (2CH); 127.68 (Q); 126.96 (2CH)
and 126.94 (2CH), incompletely resolved; 124.07 (2CH);
64.53 (C6H5C); 58.21 (NCMe3); 44.80 (bridgehead CH);
27.29 (methyls).
ϭ
ϭ
the impurity acetoxypyrrolidinedione appeared to be
present: 5.39 (1H, dd, 3J
ϭ 8.78 and 5.04, HX ); 3.13 (1H,
2
2
dd, J
ϭ
18.26, 3J
ϭ
8.78, HM); 2.69 (1H, dd, J
ϭ 18.26,
N-n-Pentylamine. (1H NMR): 2.68 (2H, t, J
NCH2); 1.44 (2H, quintet, J
(4H, m, N(CH2)2CH2CH2); 1.22 (2H, br s, NH2); 0.90
(3H, t, J 6.88, Me).
N-(n-Pentyl)maleamic Acid. Mp 64–68
72.5Њ.12 (1H NMR): 7.92 (1H, br s, NH); 6.50 (1H, slight-
ly br d, J 12.77, half of AB q, HC CH); 6.32 (1H, d,
12.77, half of AB q, HC CH); 3.37 (2H, q, J
6.78, NCH2); 1.61 (2H, quintet, J 7.22, NCH2CH2);
ϭ
6.97,
3J
ϭ
5.03, HA); 2.10 or 2.17 (tentative, 3H, s, AcO). None
ϭ
6.94, NCH2CH2); 1.32
of the precursor maleamic acid remained based on the
absence of its AB q signals, but traces (ca. 5%) of a trans
ϭ
AB q system were seen: 6.97 (1H, d, J
ϭ 15.35); 6.82
Њ
, lit. mp 72Њ 11
;
(1H, d, J 15.31). This crude maleimide was satisfactory
ϭ
for subsequent direct reaction with phencyclone.
ϭ
ϭ
N-(1-Ethylpropyl)maleimide Adduct with Phency-
J
ϭ
ϭ
ϭ
clone. Recrystallized twice. Mpr 236–240
Њ
(dec.) IR
(polyethylene card, cmϪ ): 1789.9 (strained ketone);
1702.4 (NC O); 1447.4; 1396.3; 1358.1; 1202.4; 753.1;
724.2; 697.3. (13C NMR): 196.47 (ketone); 174.91
(NC O); 133.90 (Q); 133.47 (Q); 131.52 (Q); 130.96
1
ϭ
1.33 (4H, m, N(CH2)2CH2CH2); 0.90 (3H, t, J
Me).
ϭ
6.77,
ϭ
N-(n-Pentyl)maleimide. Obtained as oil; lit. mp: oil.11
ϭ
(1H NMR): 6.69 (2H, s, HC
NCH2); 1.58 (2H, quintet, J
ϭ
CH); 3.51 (2H, t, J
ϭ
7.33,
(2CH); 129.28 (2CH); 129.20 (2CH); 128.54 (2CH);
128.33 (2CH); 126.97 (2CH); 126.53 (2CH); 126.36 (Q);
125.87 (2CH); 122.85 (2CH); 63.43 (C6H5C); 56.81
(NCH2); 44.09 (bridgehead); 23.47 (NCH2CH2); 10.05
(Me). See Discussion section regarding unusual broad-
ening of methyl signal in proton NMR.
ϭ
7.39, NCH2CH2); 1.30
(4H, complex m, N(CH2)2CH2CH2); 0.89 (3H, t, J
Me).
ϭ
7.04,
N-(n-Pentyl)maleimide Adduct with Phencyclone.
Mpr 240–245
(dec.) (13C NMR): 196.80 (ketone);
174.27 (NC O); 133.85 (Q); 133.31 (Q); 131.27 (Q);
Њ
ϭ
t-Amylamine. (1H NMR): 1.388 (2H, q, J
CH2);1.20 (2H, very broad, NH2); 1.072 (6H, s, gem
methyls); 0.893 (3H, t, J 7.51, CH3).
N-(t-Amyl)maleamic Acid. Mpr 91–105
8.09 (1H, br s, NH); 6.48 (1H, d, J 12.86, half of AB
q, HC CH); 6.23 (1H, d, J 12.92, half of AB q,
HC CH); 1.78 (2H, q, J
methyls); 0.869 (3H, t, J
ϭ 7.51,
130.91 (2CH); 129.33 (2CH); 129.12 (2CH); 128.60
(2CH); 128.37 (2CH); 127.01 (2CH); 126.60 (2CH);
126.35 (Q); 125.83 (2CH); 122.92 (2CH); 63.35 (C6H5C);
44.43 (bridgehead); 38.71 (NCH2); 27.92; 26.06; 21.57;
13.30.
ϭ
Њ
. (1H NMR):
ϭ
ϭ
ϭ
N-(i-Pentyl)maleamic Acid. Mpr 110–120
NMR): 8.21 (very br, s, 1H, NH); 6.54 (1H, d, J
half of AB q, HC CH, unusual ‘‘slope-topped’’ peaks,
see Discussion section); 6.30 (1H, d, J 12.80, half of
AB q, HC CH); 3.39 (2H, approx. q [symmetrical six
lines], J ca. 7, NCH2); 1.65 (1H, nonet, J 6.64,
CHMe2); 1.49 (2H, q, J 7.26, NCH2CH2);) 0.93 (6H,
d, J 6.57, methyl).
N-(i-Pentyl)maleimide. Oil. (1H NMR): 6.68 (2H, s,
HC CH); 3.53 (2H, approx. t, J ca. 7.39, NCH2); 1.55
(1H, m, CH); 1.46 (2H, m, NCH2CH2); 0.93 (6H, d, J
6.38, Me2). About 25% of the impurity acetoxypyrroli-
dinedione appeared to be present: 5.42 (1H, dd, 3J
8.65
8.67, HM);
4.77, HA); 2.10 or 2.17
Њ
. (1H
12.83,
ϭ
ϭ
7.49, CH2); 1.36 (6H, s, gem
ϭ
ϭ
7.48, CH3). See Discussion
ϭ
section for unusual peak shapes of AB q signals and time-
dependence of spectra.
ϭ
ϭ
N-(t-Amyl)maleimide. Oil. (1H NMR): 6.53 (2H, s,
ϭ
HC
methyls); 0.81 (3H, t, J
appeared to contain ca. 25% acetoxypyrrolidinedione im-
ϭ
CH); 1.90 (2H, q, J
ϭ
7.42, CH2); 1.57 (6H, s, gem
ϭ
ϭ 7.44, CH3). The crude material
ϭ
3
purity: 5.34 (1H, dd, J
ϭ 9.20 and 5.52, HX ); 3.05 (1H,
2
2
ϭ
dd, J
3J
5.51, HA); 2.16 (3H, s, AcO). A possible trace (ca.
2%) of a (trans) fumaramic acid impurity may have been
present: 6.87 (1H, d, J 15.44, half of AB q); 6.74 (1H,
d, J 15.44, half of AB q) suggesting the trans HC CH
moiety.
N-(t-Amyl)maleimide Adduct with Phencyclone.
The crude adduct was recrystallized. Mpr 236–240
(dec.) (13C NMR, CDCl3): 196.93 (ketone); 175.72
(NC O); 133.99 (Q); 133.51 (Q); 131.31 (Q); 130.95
(2CH); 129.29 (2CH); 129.10 (2CH); 128.56 (2CH);
ϭ ϭ 9.19, HM); 2.62 (1H, dd, J ϭ 18.02,
18.02, 3J
ϭ
ϭ
ϭ
ϭ
2
3
and 4.77, HX ); 3.15 (1H, dd, J
ϭ
18.32, J
ϭ
ϭ
ϭ
2
3
2.66 (1H, dd, J
(tentative, 3H, s, AcO).
N-(i-Pentyl)maleimide Adduct with Phencyclone.
ϭ 18.34, J ϭ
Њ
1
Mpr 276–279
1792.0 (strained ketone); 1703.1 (NC
1447.5; 1395.6; 1367.2; 1187.5; 1139.3; 778.8; 754.2;
Њ
(dec). IR (polyethylene card, cmϪ ):
ϭ
O); 1499.0;
ϭ
APPLIED SPECTROSCOPY
357

128.29 (2CH); 126.97 (2CH); 126.54 (overlap of Q and
2CH); 125.94 (2CH); 122.93 (2CH); 63.67 and 62.16
(CMe2 and C6H5C); 43.77 (bridgehead CH); 31.44 (CH2);
24.93 (gem methyls); 7.99 (methyl). (13C NMR, ca. 1:1
(v/v) CDCl3:d6-acetone, shifts referenced to CDCl3):
of oxygen present and thus potentially suppress the un-
desired side-reaction of oxidative decarbonylation of
phencyclone to 9,10-dibenzoylphenanthrene.3 Most of the
crude maleimides indicated, based on their proton NMR,
the presence of significant amounts of compounds 6, for-
mally corresponding to the acetic acid addition products
from the maleimides. These impurities were ascribed to
the acetoxypyrrolidinediones,10 primarily based on the
presence of three double doublet (dd) absorptions in the
spectra. Details of these signals are included in the list-
ings of maleimide 1H NMR data in the Experimental sec-
tion for several of the compounds, 2. The compounds 6
would also be expected to show a singlet ca. 2.17 ppm
for the acetoxy methyl, as well as signals for the N-alkyl
moieties; we have not attempted to report the latter sig-
nals. We also observed in the 1H NMR spectra for several
crude maleimides small amounts of AB q signals, ca.
6.5–7.0 ppm, with large apparent splittings of ca. 15.5
196.09 (ketone); 175.05 (NCϭO); 133.67 (Q); 132.93
(Q); 130.52 (Q); 130.28 (2CH); 128.54 (2CH); 128.23
(2CH); 127.44 (2CH); 127.26 (2CH); 126.08 (2CH);
125.85 (Q); 125.43 (2CH); 125.25 (2CH); 122.21 (2CH);
62.83 and 60.76 (CMe2 and C6H5C); 43.07 (bridgehead
CH); 30.56 (CH2); 23.98 (gem methyls); 6.92 (methyl).
N-(1,1,3,3-Tetramethyl)butyl Amine (‘‘t-Octyl’’
Amine). (1H NMR): 1.44 (2H, s, CH2); 1.31 (2H, v br s,
NH2); 1.18 (6H, s, Me2); 1.02 (9H, s, Me3).
N-(1,1,3,3-Tetramethyl)butyl Maleamic Acid (‘‘t-
Octyl’’ Maleamic Acid). Mp 125–132
139
(10). (1H NMR): 7.40 (1H, br s, NH); 6.30 (1H, d,
12.87, half of AB q, HC CH); 6.24 (1H, d, J
12.86, half of AB q, HC CH); 1.82 (2H, s, CH2); 1.47
Њ, lit. mp 138–
Њ
J
ϭ
ϭ
ϭ
ϭ
Hz, consistent with the trans substituted –HCϭCH–
(6H, s, Me2); 1.00 (9H, s, Me3).
group,13,14 which may represent some isomerization of the
precursor cis maleamic acids (3J ca. 12.8 Hz) to the ‘‘fu-
maramic acids’’ under the cyclodehydration conditions.
[We refer to ‘‘AB q’’ patterns in describing the AB pat-
terns of two coupled doublet signals, commonly (but im-
precisely) called ‘‘quartets.’’] The presence of impurities
in our maleimide samples had little effect on our ability
to isolate clean target phencyclone adducts. For the pro-
foundly hindered 1,1,3,3-tetramethylbutyl system (‘‘t-oc-
tyl’’) of 5h and 2h, we employed the alternative method
of Wang for preparation of the maleamic acid and mal-
eimide,10 in which the cyclodehydration step is achieved
with acetic anhydride and triethylamine in refluxing ac-
etone.
The main goal for NMR characterizations of the ad-
ducts was to determine if the unsubstituted bridgehead
phenyl groups were undergoing slow rotation as indicated
by slow exchange limit (SEL) proton and carbon-13
spectra. For adducts in the SEL regime, nine pairs of aryl
methine signals are expected (barring coincidental over-
N-(1,1,3,3-Tetramethyl)butyl Maleimide (‘‘t-Octyl’’
Maleimide). Prepared according to Wang.10 Mp 59.5–
60Њ ϭCH);
, lit. mp 59–60Њ.10 (1H NMR): 6.52 (2H, s, HC
1.89 (2H, s, CH2); 1.66 (6H, s, Me2); 0.92 (9H, s, Me3).
N-(1,1,3,3-Tetramethyl)butyl Maleimide (‘‘t-Octyl’’
Maleimide) Adduct with Phencyclone. Recrystallized,
1
mpr 276–278
Њ
(dec). IR (polyethylene card, cmϪ ):
1792.4 (strained ketone); 1701.5 (NC
ϭ
O); 1549.3;
1497.2; 1477.1; 1448.1; 1366.4; 1334.6; 1250.4; 1223.0;
1186.2; 1140.1; 781.9; 757.5; 724.7; 699.5. (13C NMR):
196.96 (ketone); 175.81 (NCϭO); 133.99 (Q); 133.53
(Q); 131.25 (Q); 130.95 (2CH); 129.30 (2CH); 129.12
(2CH); 128.58 (2CH); 128.31 (2CH); 126.98 (2CH);
126.585 (Q) and 126.558 (2CH) overlapped; 125.97
(2CH); 122.99 (2CH); 63.70 and 62.58 (C6H5C and NC);
49.73 (CH2); 43.69 (bridgeheads); 31.29 (CMe3); 30.94
(Me3); 27.67 (Me2).
RESULTS AND DISCUSSION
laps), including four pairs (4
threnoid moiety and five pairs (5
ϫ
2CH) for the phenan-
2CH) from the
Compound structures, atom numbering, and synthetic
pathways are shown in Fig. 1. In general, the desired
alkyl groups were derived from the corresponding pri-
mary amines, 4, which reacted with maleic anhydride to
form the N-alkylmaleamic acids, 5. The maleamic acids
usually underwent cyclodehydration in acetic anhydride
ϫ
bridgehead phenyls. Four pairs of non-protonated aryl
quaternary (Q) carbon signals are expected in the C-13
NMR (assuming no coincidental overlaps), including
three from the phenanthrenoid system and one from the
ipso C1Ј of the phenyls. With the exception of adduct 3c,
with sodium acetate, ca. 100Њ, to form the crude maleim-
spectra of aryl carbons in CDCl3 provided nine distinct
stronger signals for the pairs of aryl methines and four
weaker signals for the quaternary aryl signals, consistent
with SEL systems (on the carbon NMR timescale).
CD2Cl2 was a helpful solvent for adduct 3c for improved
aryl carbon signal dispersion. For this adduct, benzene
had been used as solvent for reaction of 1 with 2c. It has
been reported that phencyclone adducts often form com-
plexes (solvates) with various solvents15–17 and our ini-
tially isolated sample of 3c revealed a strong benzene
peak in the 13C NMR, overlapped with adduct signals.
Removal of the interfering benzene was achieved by sev-
eral cycles of dissolution of crude adduct in CH2Cl2 fol-
lowed by solvent removal. These ‘‘codistillations’’ ap-
peared to effectively remove interfering benzene NMR
signals. Use of a relaxation delay (RD) of 3 s for routine
13C NMR resulted in facile distinction of protonated and
ides, 2, albeit in low to moderate yields. An excess of
the crude maleimides resulted in decolorization of the
intense green-black phencyclone, 1, to indicate the con-
sumption of the phencyclone. The virtue of these synthe-
ses is that the crude 2 can be used with essentially no
purification, in moderate excess. The lower N-alkylmal-
eimides examined here, and their common impurities (see
below), have rather low melting points and are quite sol-
uble; the excess maleimides and impurities are readily
separated from the phencyclone adducts, 3, which have
very low solubilities and high melting/decomposition
points. The key target adducts, therefore, are easily ob-
tained in good purity based on proton NMR spectra. Most
of the phencyclone adducts were obtained by reaction of
the maleimide in a sealed screw-cap vial at ambient tem-
perature, with minimal headspace to reduce the amount
358
Volume 59, Number 3, 2005

non-protonated carbon signals, with the latter being much
weaker; the protonated carbon signals had areas roughly
in proportion to their numbers.
1
For the 1D H NMR spectra of the adducts in CDCl3,
several overlaps occurred for aryl proton resonances for
each adduct. Dispersing the signals into a second dimen-
sion with the high-resolution COSY45 spectra permitted
explicit recognition of connectivities in the (CH)5 spin
system from the bridgehead phenyls (demonstrating slow
1
phenyl rotation on the H NMR timescale) as well as the
(CH)4 spin system from the phenanthrenoid moiety. Un-
der our conditions, aryl proton crosspeaks corresponding
3
not only to vicinal J couplings, but also to long-range 4J
5
and J couplings, were readily observable. Approximate
proton shifts (and gross signal multiplicities) could be
obtained from the COSY spectra even when overlaps
were seen in the 1D spectra. Representative spectra are
presented in Fig. 2.
For the adduct 3f, with the 1-ethylpropyl alkyl group,
1
an unusual feature was seen in the H NMR. The ex-
pected 6H intensity triplet signal for the two methyl
groups appeared at
Ϫ0.02 ppm (upfield of TMS) excep-
tionally broadened, totally lacking in the normal valleys
between the branches of the triplet, showing only slight
inflection points. (Slight broadening was discernable in
the complex multiplet assigned to the CH2 signal.) Other
signals in the spectrum were sharp, suggesting the pos-
sibility of signal broadening from a dynamic exchange
process.18 We suggest that the broadened signals may re-
flect hindered rotations within the N–CH(CH2CH3)2 moi-
ety. For example, if a favored conformation has the two
ethyl groups in different orientations, experiencing dif-
fering anisotropic shielding from the phenanthrenoid
group, substantial chemical shift differences (⌬␯) may be
associated with each different spatial orientation. With
rapid bond rotations interconverting the possible con-
formers, a sharply averaged fast exchange limit (FEL)
spectrum could be expected, producing a single sharply
averaged triplet for the methyl resonance. If the bond
rotations were exceedingly slow on the NMR timescale,
two separate triplet resonances might be observable as an
SEL spectrum, reflecting a pair of methyl groups in two
different environments. With intermediate rates of the ex-
change process, signal broadening might result. For the
1-ethylpropyl group, rotation about the N-alkyl bond may
be the critical process, with hindrance resulting from the
imide carbonyls and the two ethyl groups of the N-alkyl
moiety. We concur with a referee’s suggestion that future
variable temperature NMR studies would potentially be
quite informative here, e.g., to see if elevated tempera-
tures provide the anticipated simplified FEL spectrum.
(The 1-ethylpropyl case may be especially complex since
the two hydrogens of each CH2 group are formally dia-
←
F
IG. 2. Selected spectral expansions from 1H NMR spectra (300 MHz),
in CDCl3 at ambient temperatures. Chemical shifts are in ppm. Traces
(a) and (b) for the phencyclone adduct of N-(1-ethyl)propylmaleimide,
showing: (a) slight broadening in the signal for the CH2 groups, spectral
window ca. 0.75 to 1.20 ppm, and (b) severe broadening in the (ex-
eimide, spectral window ca.
near 2.92 ppm is not shown.) The 1H multiplet of the gamma
NCH2CH2CH(CH3)2 appears at 0.24 ppm, the 6H doublet of the methyls
at 0.16 ppm, and the highly shielded beta NCH2CH2 is seen at
ppm as a 2H gross quartet. The lower trace shows the aryl proton region
for the isoamylmaleimide adduct. Aryl proton assignments are as in-
dicated. (Note that different scales are used on the upper and lower
traces.)
Ϫ0.5 to 0.5 ppm. (The NCH2 triplet signal
Ϫ0.41
pected triplet) signal for the methyls, spectral window ca.
Ϫ0.16 to 0.12
ppm. Note that the methyl signal is highly shielded, appearing upfield
of the (sharp) TMS peak at 0.0 ppm. In part (c), the upper trace shows
the upfield portion of the phencyclone adduct of N-isoamylmal-
APPLIED SPECTROSCOPY
359

stereotopic, regardless of conformation, and this
CH(a)H(b) system could potentially result in considerable
spectral complexity, depending on the ⌬␯ magnitudes.)
Some suggestion of unusual dynamic NMR effects from
possible intermediate exchange rates may have been seen
1
in the H NMR (in CDCl3) of maleamic acids 5e (N-
isoamylmaleamic acid) and 5g (N-t-amylmaleamic acid).
For the expected four-line AB q portion of the spectrum
from the HCϭCH moiety, 5e displayed a sharp doublet
for the highfield half, but the lowfield half had peaks that
were somewhat flat-topped. In contrast, the AB q portion
of the spectrum of 5g had a sharp pair of peaks for the
lowfield half, but distinctly skewed triangle-shaped peaks
for the highfield pair of lines. We suggest that the ma-
leamic acids can be quite complex in terms of their var-
ious possible conformers, which would include not only
rotamers resulting from rotations about the N-alkyl NC
bond, but also from rotations about the amide HN–CO
bonds.19,20
Several features of the 1H NMR spectra of the adducts
are noteworthy with respect to apparent magnetic aniso-
tropic shielding of the N-alkyl groups. These shielding
effects are consistent with normal endo stereochemistry
for the Diels-Alder adducts,15 causing the N-alkyl groups
to occupy space proximal to the phenanthrenoid moiety.
The proton shifts for the adducts are presented in Table
I. Comparison of chemical shifts for protons in the alkyl
groups of the adducts, 3, with the shifts for corresponding
protons in the maleimides, 2, may be taken as an ap-
proximation of the anisotropic effects from the phenan-
threnoid group; other groups in the adducts, e.g., the
bridgehead phenyls, may also have some anisotropic in-
fluences. Since the upfield (alkyl) portions of the proton
NMR spectra of the adducts 3 appear to exhibit a single
set of sharp signals (with the possible exception of 3f,
discussed above), the alkyl portions must be rotating rap-
idly on the NMR timescale to give sharply averaged FEL
spectra, or the alkyl groups occupy only (or predomi-
nantly) a single conformation. Simple C–C bond rota-
tions within the alkyl groups should be fast at ambient
temperatures, so we generally favor rapid interconver-
sions of different alkyl group conformers. The observed
chemical shifts exhibited by the alkyl group protons
would then be the time-averaged shifts of the protons in
each conformation, weighted according to the conform-
er’s mole fraction. In order to obtain a better understand-
ing of possible adduct conformer structures, we have ap-
plied computational methods for geometry optimizations
of the adducts (see below.)
Table II presents the magnetic anisotropic shielding
magnitudes for protons in the alkyl moieties of the phen-
cyclone adducts, calculated as the differences in chemical
shifts for the protons in the maleimides and the shifts of
the corresponding protons in the adducts, i.e.,
␦ (maleim-
ides) Ϫ ␦ (adducts). We feel that these shielding magni-
tudes would largely be attributable to phenanthrenoid an-
isotropic shielding effects for the endo adducts, but fur-
ther contributions could result from changes in the con-
former populations of the N-alkyl groups for maleimide
versus adduct. In general, shielding is greatest for beta
protons, NCCH, but the magnitudes vary considerably,
from as little as 0.63 ppm for the CH2 in the t-amyl sys-
tem, to as much as 1.87 ppm in the n-pentyl and isopentyl
360
Volume 59, Number 3, 2005

TABLE II. Anisotropic shielding magnitudes, ⌬␦ (ppm), for alkyl
protons in adducts relative to maleimides: chemical shift (maleim-
n-pentyl, we show the results for a syn and an anti en-
ergy-minimized conformer. In each system, the n-alkyl
chain is fully extended (‘‘zig-zag’’) with the syn and anti
ide)
Ϫ
chemical shift (adduct).
Nuclei Maleimide
NCH
Adduct
⌬␦ (ppm)
isomers essentially interconverted by a ca. 180Њ rotation
Isopropyl
Isobutyl
4.34
1.39
3.34
2.02
0.89
1.58
3.51
1.58
ca. 1.30
ca. 1.30
0.89
3.53
1.46
1.55
0.93
3.83
1.70
0.82
1.57
1.90
0.81
1.66
1.89
0.92
3.62
0.34
2.74
0.66
0.72
1.05
0.60
1.36
0.99
1.01
0.65
1.87
1.00
0.72
0.49
0.61
1.87
1.31
0.77
0.56
0.70
0.84
1.08
0.63
0.61
1.18
0.72
0.22
about the N–CH2 bond. In the syn structures, the alkyl
chain is much closer to the phenanthrenoid moiety central
ring where larger anisotropic shielding magnitudes might
be experienced. Unexpectedly, the syn structures were
found to be lower in energy than the trans. For the t-
butyl, t-amyl and ‘‘t-octyl’’ systems, the conformers la-
beled syn each had an alkyl group R, i.e., N–C(alpha)–
R, oriented roughly perpendicular to the imide plane, and
directed ‘‘in’’ toward the phenanthrenoid moiety; these R
groups corresponded to methyl, ethyl, and ‘‘neopentyl’’
[CH2C(Me)3], respectively, for the three adducts, 3c, 3g,
Me2
NCH2
CH
Me2
Me3
Ϫ
0.10
0.57
2.86
0.29
0.30
0.58
0.40
2.92
0.41
0.24
0.16
3.27
1.00
0.02
0.49
1.27
0.20
0.48
1.17
0.70
t-Butyl
n-Pentyl
NCH2
CH2 (
CH2 (
CH2 (
Me
NCH2
CH2 (
␤
)
Ϫ
␥
)
␦
)
Isopentyl
␤
␥)
)
Ϫ
Ϫ
CH (
Me2
and 3h. With ca. 180Њ rotation about the N–C(alpha)
bond, the corresponding anti conformers are formally
generated. Note that the anti conformers were the more
stable for the t-butyl and t-amyl systems, but the syn was
very slightly more stable for the t-octyl. For the isobutyl
and isoamyl adducts, 3b and 3e, the respective syn con-
formers each had their N–CH2(alpha)–R groups approx-
imately oriented perpendicular to the imide plane and di-
rected ‘‘in’’ toward the phenanthrenoid moiety, with the
N–CH2(alpha) protons angled outwards and away from
the phenanthrenoid. The corresponding anti conformers
1-Ethylpropyl NCH
CH2
Me
t-Amyl
N-CMe2
N-CCH2
N-CCH2Me
N-CMe2
CH2
‘‘t-Octyl’’
CMe3
cases. For the alpha protons, NCH, shielding magnitudes
vary over a narrow range, from 0.56 ppm in 1-ethylpro-
pyl to 0.72 ppm in isopropyl. At the gamma position,
NCCCH, the shielding magnitudes were as little as 0.61
ppm for t-amyl up to 1.31 ppm in isopentyl. The 1-ethyl-
propyl adduct was unique among the adducts in exhib-
iting a larger shielding magnitude at the gamma position,
for the methyls (0.84 ppm), than for the beta protons
(0.70 ppm). All other adducts showed increasing shield-
ing magnitudes going from the alpha to the beta, and
then decreasing magnitudes from the beta to the gamma
position. Since the phenanthrenoid moiety presumably
causes hindrance on one face of the pyrrolidinedione por-
tion of the adducts, it could influence the relative amounts
of differing rotational conformers of the N-alkyl groups,
and produce apparent changes in the shielding magni-
tudes.
Ab Initio Structure Calculations. Geometry optimi-
zation calculations were performed at the Hartree-Fock
level with the 6-31G* basis set. This is considered to
provide good accuracy for structures and energies of con-
formers at reasonable cost in terms of computation
times.21–24 The selected software and hardware permit cal-
culations with the large numbers of basis functions for
these adduct molecules. Our present ab initio HF/6-31G*
calculations appear to be the highest level that has been
applied to these compounds. For most of the adducts, we
have obtained results for two different conformers, with
key differences associated with orientations in the N-al-
kyl moieties. (The N-isopropylmaleimide adduct seemed
to give only a single minimized structure, with the NCH
methine proton essentially coplanar with the imide, and
with one of the methyls angled ‘‘in’’ toward the phen-
anthrenoid and the other methyl angled ‘‘out.’’) We have
also included the N-n-butylmaleimide adduct in the pres-
ent calculations; NMR data were published earlier.2 For
both of the N-n-alkylmaleimide systems, i.e., n-butyl and
were formally related by a rotation of ca. 180Њ rotation
about the N–C(alpha) bond. The more stable conformer
was the anti for the isobutyl adduct, 3b, but the syn was
more stable for the isoamyl adduct, 3e. Perhaps the most
complex conformer pair was found for the 1-ethylpropyl
system, adduct 3f. The two minimized structures that
were obtained were not approximated (interconverted) by
the 180Њ rotation about the N–C(alpha) bond that inter-
related the conformer pairs described above. For the con-
former of 3f which we have arbitrarily labeled ‘‘syn’’ (the
more stable of the two conformers), a dihedral angle of
5.01
for O
Њ
was seen for the O
ϭC–N–C–H moiety, and 121.61Њ
ϭ
C–N–C–C(beta, ‘‘in’’). [C(beta, ‘‘in’’) refers to
the beta carbon in the ethyl group directed into the adduct
cavity; the same carbonyl group is used in measuring
both dihedral angles.] For the corresponding ‘‘anti’’ con-
former of 3f, the respective dihedral angles were 16.58
Ϫ
Њ
for the O C–N–C–H moiety, and 97.07 for OϭC–N–
ϭ
Њ
C–C(beta, ‘‘in’’). Selected calculated structures are
shown in Fig. 3.
Table III presents some of the calculated parameters
for these conformers. The data include the energies for
the optimized conformer structures in atomic units (har-
trees) [1 hartree
the energy differences,
ϭ
627.5 kcal/mol]. For conformer pairs,
E (kcal/mol), are shown to give
⌬
a sense of relative equilibrium abundances. Additional
conformers may also contribute significantly but we have
limited the present treatment to just the two tabulated
structures of a given alkyl group. For the systems listed,
the dihedral angles for C2ЈC1ЈCCϭO are indicated, as
measured to each phenyl group in the adduct structure.
The dihedral angles of the bridgehead phenyl groups are
defined with C2Ј arbitrarily designated as proximal to the
bridging ketone carbonyl. Generally, the dihedral angle
magnitudes for the pair of phenyls in each structure are
in close agreement, and all of the adducts have calculated
dihedrals of ca. 51.5–52.5
Њ
. Note that the signs of the
APPLIED SPECTROSCOPY
361

FIG. 3. Representative views of calculated structures, HF/6-31G*, for phencyclone adducts. The hydrogens are ‘‘hidden’’ for clarity. (a) compound
3a, from N-isopropylmaleimide; (b) adduct from N-n-butylmaleimide, syn at left, anti at right; (c) compound 3c, from N-t-butylmaleimide, syn at
left, anti at right; (d) compound 3e, from N-isoamylmaleimide, syn at left, anti at right; (e) compound 3f, from N-(1-ethylpropyl)maleimide, syn
at left, anti at right; (f) compound 3g, from N-t-amylmaleimide, syn at left, anti at right; and (g) compound 3h, from N-(1,1,3,3-tetrame-
thyl)butylmaleimide, syn at left, anti at right.
362
Volume 59, Number 3, 2005

TABLE III. Calculated parameters for branched and n-alkyl adducts. See Notes and Discussion. Angles are given in degrees; Pucker and
N-pyramidalization are expressed in angstroms.a
Dihedral angles
C2 C1 CC
Pucker
(Angstroms)
N-pyram.
N in/out
Energy (AU)
Ј
Ј
ϭ
O
Isopropyl
Ϫ1658.0944057
52.04,
Ϫ51.84
0.277
0.006 (out)
n-Butyl
‘‘syn’’
Ϫ
1697.1301953
(0.243)
51.79,
51.68,
Ϫ
51.81
51.76
0.275
0.273
0.011 (out)
0.002 (out)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1697.1298077
Ϫ
Isobutyl
‘‘syn’’
Ϫ
1697.1282848
(0.513)
52.40,
51.95,
Ϫ
50.95
51.82
0.285
0.276
0.041 (in)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1697.1291027
Ϫ
0.034 (out)
t-Butyl
‘‘syn’’
Ϫ
1697.1167705
(0.153)
52.30,
52.18,
Ϫ
52.17
52.30
0.276
0.270
0.085 (out)
0.078 (in)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1697.1170137
Ϫ
n-Pentyl
‘‘syn’’
Ϫ
1736.1649380
(0.236)
51.80,
51.74,
Ϫ
51.87
51.84
0.278
0.272
0.011 (out)
0.003 (out)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1736.1645613
Ϫ
Isopentyl
‘‘syn’’
Ϫ
1736.1634634
(0.257)
51.86,
51.93,
Ϫ
51.93
51.46
0.278
0.273
0.011 (out)
0.003 (out)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1736.1630538
Ϫ
1-Ethylpropyl
‘‘syn’’
Ϫ
1736.1571675
(0.294)
52.45,
52.29,
Ϫ
51.68
51.42
0.274
0.268
0.003 (in)
0.003 (in)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1736.1566989
Ϫ
t-Amyl
‘‘syn’’
Ϫ
1736.1481935
(0.148)
52.23,
52.26,
Ϫ
52.25
52.24
0.275
0.272
0.083 (out)
0.077 (in)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1736.1484295
Ϫ
‘‘t-Octyl’’
‘‘syn’’
Ϫ
1853.2392203
(0.047)
51.99,
52.44,
Ϫ
52.45
52.30
0.284
0.271
0.083 (out)
0.079 (in)
⌬E (kcal/mol)
‘‘anti’’
Ϫ1853.2391447
Ϫ
aCalculations were performed at the Hartree-Fock level with the 6-31G* basis set. Pucker is defined as the distance between the midpoint of the
line joining H2 and H7 and the center of the C4a–C4b bond. N-Pyramidalization is measured as the distance of the nitrogen from the plane defined
by the three directly bonded atoms. If the nitrogen is ‘‘in,’’ it lies on the side of the plane closer to the phenanthrenoid moiety.
tabulated dihedral angles reflect the sense of the axial
chirality in selected portions of the structures, as opti-
mized. Data in Table III indicate considerable similarity
in the anti and syn conformers with respect to the aro-
matic ring portions of the structures. Puckering (nonplan-
arity) of the phenanthrenoid moiety is quantitatively ex-
pressed as the distance between the midpoint of the line
joining H2 and H7 and the center of the C4a–C4b bond.
If this three-ring aromatic portion is regarded as having
a ‘‘butterfly-like’’ structure, then the outer rings
(‘‘wings’’) are slightly folded toward the N-alkyl chain;
calculated distances for all structures were ca. 0.27–0.28
these three compounds, the nitrogen has moved to op-
posite sides of the plane of its directly bonded atoms.
Further geometric parameters (distances and defined
angles) are summarized in Table IV for several of the
geometry-optimized structures. Selected distances from
specified atoms to defined points in the phenanthrenoid
groups are used to characterize the structures. ‘‘Point 1’’
was defined as the midpoint of the line joining C8a and
C10a, and was regarded as the nominal ‘‘center’’ of the
middle ring of the phenanthrenoid system. ‘‘Point 2’’ was
the midpoint of the bond joining C9 and C10. We have
tabulated not only the distances from each of the hydro-
gens of the alkyl groups to Point 1, but also the angles
defined by: Point 2, to Point 1, to Hydrogen. These angles
provide a rough measure of how close the alkyl protons
are (with respect to angular measure) to the approximate
perpendicular axis of the central phenanthrenoid ring, i.e.,
˚
A. Pyramidalization at nitrogen is measured as the dis-
tance of the nitrogen from the plane defined by the three
atoms that are directly bonded to the nitrogen. We have
indicated in Table III whether the nitrogen is ‘‘in’’, i.e.,
located on the side of the plane toward the phenanthren-
oid moiety, or ‘‘out’’, on the side of the plane remote
from the phenanthrenoid portion. For the most hindered
structures, i.e., t-butyl, t-amyl, and 1,1,3,3-tetramethyl-
butyl (‘‘t-octyl’’), the greatest N-pyramidalization was
measured. Note that for the syn/anti conformer pairs of
an angle of 90Њ would indicate that the proton lies close
to the axis, where it should receive maximum anisotropic
shielding with respect to the angular part of this shield-
ing. This approximation should be most valid for n-al-
kylmaleimide adducts, since the carbons of the n-alkyl
APPLIED SPECTROSCOPY
363

TABLE IV. Geometric parameters for branched smaller alkyl systems. Distances (angstroms) and angles (degrees) to specified protons in
alkyl group. See Notes and Discussion.a
Isopropyl
‘‘anti’’
CH (alpha)
4.985
77.13
Me ‘‘in’’
3.544, 4.379, 5.264
100.93, 91.47, 98.37
Me ’’out’’
6.532, 6.152, 6.767
63.28, 70.09, 78.11
Dist.
Angles
n-Butyl
‘‘syn’’
H(alpha)
H(beta)
H(gamma)
H(delta)
5.585, 5.505, 6.952
121.03, 120.98, 116.82
8.803, 8.803, 9.272
72.87, 72.87, 81.88
Dist.
Angles
Dist.
5.382, 5.383
73.42, 73.41
4.551, 4.545
83.03, 83.05
3.785, 3.787
97.61, 97.59
6.407, 6.405
65.50, 65.50
6.245, 6.247
97.46, 97.44
7.073, 7.072
86.07, 86.07
‘‘anti’’
Angles
t-Butyl
Me ‘‘in’’
Me ‘‘out’’ (1)
Me ‘‘out’’ (2)
‘‘syn’’
(one Me ‘‘in’’)
Dist.
Angles
3.894, 3.894, 5.306
95.51, 95.48, 97.36
Me ‘‘in’’ (1)
5.699, 6.454, 6.715
75.63, 63.89, 78.70
Me ‘‘in’’ (2)
5.701, 6.454, 6.716
75.58, 63.86, 78.68
Me ‘‘out’’
‘‘anti’’
(one Me ‘‘out’’)
Dist.
Angles
3.823, 5.149, 5.448
96.79, 82.71, 96.19
3.819 5.145, 5.445
96.83, 82.75, 96.22
6.326, 6.325, 6.744
65.90, 65.90, 78.36
t-Amyl
CH2
CH2Me ‘‘in’’
CMe (1)
CMe (2)
‘‘syn’’
(ethyl ‘‘in’’)
Dist.
Angles
3.812, 3.812
94.43, 94.41
CMe ‘‘in’’ (1)
3.800, 5.186, 5.409
95.80, 82.58, 96.44
5.535, 6.260, 6.259
112.17, 98.78, 98.77
CMe ‘‘in’’ (2)
3.797, 5.184, 5.407
95.82, 82.62, 96.47
5.718, 6.415, 6.757
75.74, 63.39, 77.95
CH2
6.263, 6.261
64.80, 64.82
5.718, 6.415, 6.756
75.70, 63.37, 77.94
CH2Me ‘‘out’’
7.324 7.323, 8.121
84.07, 84.08, 74.71
‘‘anti’’
(ethyl ‘‘out’’)
Dist.
Angles
aDistances (angstroms) are given for every proton to ‘‘Point 1,’’ the midpoint of the line joining C8a and C10a. Angles (degrees) are measured
from ‘‘Point 2’’ (the midpoint of the C9–C10 bond), to Point 1, to the specific proton.
group essentially lie on the near-mirror symmetry plane,
and the hydrogens of the n-alkyl groups lie close to the
plane which: (a) is perpendicular to the central ring of
the phenanthrenoid system, (b) bisects the bond from C-
9 to C-10, and (c) passes through Point 1 (the center of
the ring). The structures in Table IV (except for the iso-
propyl case) have calculated structures that nearly have
a mirror plane of symmetry. The optimized conformers
for the ‘‘t-octyl’’ system, i.e., N-(1,1,3,3-tetrameth-
yl)butyl, also were close to having a mirror plane of sym-
metry. In contrast, structures for isobutyl, isoamyl (iso-
pentyl), and N-(1-ethyl)propyl did not have the alkyl
moieties positioned to confer near-mirror symmetry on
the adduct structure. If the syn/anti conformer pair for
the n-butyl system is taken as an example, the Table IV
data show that the beta alkyl protons, NCH2CH2, are es-
Geometry optimizations for the phencyclone adducts
showed the effects of severe repulsions of the ortho pro-
tons, H2Ј,6Ј, of the bridgehead phenyls with: (a) the
phenanthrenoid H1,8; (b) the ketone and imide carbonyls;
and (c) the bridgehead methine CH. These interactions
must largely determine the orientations of the bridgehead
phenyls. Data in Table III indicate considerable similarity
in the anti and syn conformers with respect to the aro-
matic ring portions of the structures.
CONCLUSION
The hindered Diels-Alder adducts, 3, from phency-
clone, 1, have been prepared using a series of N-alkyl-
maleimides, with various alkyl groups ranging from iso-
propyl to t-amyl and 1,1,3,3-tetramethylbutyl. Ambient
temperature NMR spectra for CDCl3 solutions of the ad-
ducts showed slow rotation of the unsubstituted bridge-
head phenyl groups on both the proton NMR timescale
(300 MHz) and the carbon-13 NMR timescale (75 MHz),
˚
pecially close to Point 1 (3.8 A) and also lie near the
perpendicular axis through Point 1 (ca. 97.5Њ angle) for
the syn conformer. In contrast, these beta protons for the
˚
anti conformer are further from Point 1 (6.4 A) and fur-
1
13
with slow exchange limit H and C NMR spectra. Full
aryl proton assignments in the adducts were achieved us-
ing high-resolution COSY45 spectra; assignments in the
alkyl portions were also achieved. Estimates of aniso-
tropic shielding magnitudes of the alkyl groups in the
ther from the perpendicular axis through Point 1 (ca.
65.5Њ angle). For the alpha NCH2 protons, the opposite
is true, with the protons being further from Point 1 and
from the perpendicular axis in the (more stable) syn than
in the anti conformer. Since the experimental fact for N-
n-alkylmaleimide adducts is that the beta protons display
greater anisotropic shielding magnitudes than the alpha,
we consider this as supporting a predominant syn con-
former for our samples. With the observed proton NMR
spectra of these adducts exhibiting a single set of sharp
signals for the alkyl group resonances (except for the 1-
ethylpropyl system), we assume that the syn/anti con-
former interconversions are fast on the NMR timescale
(300 MHz, ambient temperature) to give FEL spectra.
The actual observed NMR chemical shifts would reflect
a time-weighted average of all of the different existing
conformers. The above qualitative treatment appears con-
sistent with observed chemical shifts in the adducts.
1
adducts were obtained by comparing H chemical shifts
for corresponding alkyl protons in the maleimides and
their phencyclone adducts. Maximum shielding magni-
tude was usually found for the alkyl protons beta to the
nitrogen, with the NCH2CH2 protons in some adducts ap-
pearing as much as 1.87 ppm to higher field than in the
maleimides, resulting in some absorptions for the beta
protons occurring to higher field than the TMS signal. Ab
initio HF/6-31G* calculations for these adducts of phen-
cyclone were performed to allow estimates of geometric
parameters in the adducts, and results indicated that the
syn conformation (of alkyl relative to phenanthrenoid)
was usually slightly lower energy than the anti. (We note
364
Volume 59, Number 3, 2005

8. M. A. Walker, J. Org. Chem. 60, 5352 (1995).
9. L. Crombie and M. Manzoor-i-Khuda, J. Chem. Soc. 2767 (1957).
10. Z. Y. Wang, Synth. Commun. 20, 1607 (1990).
11. J. R. Heitz, C. D. Anderson, and B. M. Anderson, Arch. Biochem.
Biophys. 127, 627 (1968).
that a reviewer has suggested potential future work in-
volving variable-temperature NMR studies and possible
computational studies for calculation of chemical shield-
ing tensors for specific conformations, and we are grate-
ful for these recommendations.)
12. J.-C. Dore and C. Viel, Eur. J. Med. Chem., Chim. Ther. 10, 47
(1975).
13. H. Gu¨nther, NMR Spectroscopy: Basic Principles, Concepts, and
Applications in Chemistry (John Wiley and Sons, New York, 1995),
2nd ed.
14. R. J. Abraham, J. Fisher, and P. Loftus, Introduction to NMR Spec-
troscopy (John Wiley and Sons, New York, 1988).
15. M. Eto, K. Setoguchi, A. Harada, E. Sugiyama, and K. Harano,
Tetrahedron Lett. 39, 9751 (1998).
16. Y. Yoshitake, J. Misaka, K. Setoguchi, M. Abe, T. Kawaji, M. Eto,
and K. Harano, J. Chem. Soc., Perkin Trans. 2, 1611 (2002).
17. Y. Yoshitake, J. Misaka, M. Abe, M. Yamasaki, M. Eto, and K.
Harano, Org. Biomol. Chem. 1, 1240 (2003).
ACKNOWLEDGMENTS
Partial support has been provided to R.R. by the U.S. Education
Department (Hispanic-Serving Institutions-Title V, and the Minority
Science and Engineering Improvement Program), National Science
Foundation, and the Professional Staff Congress-City University of New
York Research Award Program. Further support has been provided by
the Project ASCEND/McNair Program to R.P., N.S., and O.S. Technical
assistance was provided by Kimberly Marshall.
18. J. Sandstro¨m, Dynamic NMR Spectroscopy (Academic Press, New
York, 1982).
19. T. H. Siddall III and W. E. Stewart, J. Org. Chem. 34, 2927 (1969).
20. W. E. Stewart and T. H. Siddall III, Chem. Rev. 70, 517 (1970).
21. F. Jensen, Introduction to Computational Chemistry (John Wiley
and Sons, New York, 1999).
22. TITAN Tutorial and User’s Guide (Wavefunction, Inc., Irvine, CA,
and Schro¨dinger, Inc., Portland, OR, 1999).
23. W. J. Hehre, A. J. Shusterman, and J. E. Nelson, The Molecular
Modeling Workbook for Organic Chemistry (Wavefunction, Inc.,
Irvine, CA, 1998).
1. K. Bynum and R. Rothchild, Spectrosc. Lett. 29, 1599 (1996).
2. K. Bynum and R. Rothchild, Spectrosc. Lett. 29, 1621 (1996).
3. R. Callahan, K. Bynum, R. Prip, and R. Rothchild, Chem. Ed. 3,
S1430 (1998), http://journals.springer-ny.com/chedr.
4. O. Sklyut, R. Prip, N. Azar, R. Callahan, and R. Rothchild, Spec-
trosc. Lett. 37, 493 (2004).
5. T. Sasaki, K. Kanematsu, and K. Iizuka, J. Org. Chem. 41, 1105
(1976).
6. Y. Liwschitz, Y. Edlitz-Pfeffermann, and Y. Lapidoth, J. Am. Chem.
Soc. 78, 3069 (1956).
7. K. Takatori, T. Hasegawa, S. Nakano, J. Kitamura, and N. Kato,
Microbiol. Immunol. 29, 1237 (1985).
24. W. J. Hehre, A Guide to Molecular Mechanics and Quantum Chem-
ical Calculations (Wavefunction, Inc., Irvine, CA, 2003).
APPLIED SPECTROSCOPY
365