Investigation of Chocolate Matrix Interference on Cannabinoid Analytes

Article abstract of DOI:10.1021/acs.jafc.0c01161

The first known findings of chocolate matrix interference on cannabinoid analytes is reported. Stock solutions of four biogenic cannabinoids (Δ9-tetrahydrocannabinol, cannabidiol, cannabinol, and cannabigerol) and one synthetic cannabinoid (cannabidiol di

Full text of DOI:10.1021/acs.jafc.0c01161

Subscriber access provided by BIU Pharmacie | Faculté de Pharmacie, Université Paris V
Food and Beverage Chemistry/Biochemistry
Investigation of Chocolate Matrix Interference on Cannabinoid Analytes
David Dawson, and Robert Martin
J. Agric. Food Chem., Just Accepted Manuscript • Publication Date (Web): 01 May 2020
Downloaded from pubs.acs.org on May 2, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 20
Journal of Agricultural and Food Chemistry
Manuscript Title:
Investigation of Chocolate Matrix Interference on Cannabinoid Analytes
Authorship:
David D. Dawson, Ph.D.* (email: david.dawson@cwanalytical.com; phone: 510-545-6984)
Robert W. Martin, Ph.D.
Institution of BOTH Authors:
Cw Analytical
851 81st Ave. Ste. D
Oakland, CA 94621
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 2 of 20
Abstract:
The first known findings of chocolate matrix interference on cannabinoid analytes is
9
reported. Stock solutions of four biogenic cannabinoids (∆ -THC, CBD, CBN, CBG) and one
synthetic cannabinoid (CBDD) are subjected to milk chocolate, dark chocolate, and cocoa
powder. A clear trend of matrix interference is observed, which correlates to several chemical
factors. The amount of chocolate present is directly proportional to the degree of matrix
interference, which yields lower percent recovery rates for the cannabinoid analyte. Structural
features on the cannabinoid analytes are shown to affect matrix interference, as cannabinoids
with fewer phenolic -OH groups suffer from increased signal suppression. Additionally,
aromatization of the p-menthyl moiety appears to correlate with enhanced matrix effects from
chocolate products high in cocoa solids. These findings represent the first known documentation
of chocolate matrix interference in cannabinoid analysis, which potentially has broad implications
for complex matrix testing in the legal Cannabis industry.
Keywords:
Cannabis - matrix interference - cannabinoid - chocolate - cocoa - HPLC - matrix effects -
Cannabis testing – Cannabis-infused
ACS Paragon Plus Environment

Page 3 of 20
Journal of Agricultural and Food Chemistry
1
2
3
4
5
6
7
8
9
Introduction:
Since the legalization votes in Washington and Colorado in 2012, a total of eleven states
have legalized the recreational use of Cannabis, with another thirty-six allowing for some form of
1
medical use. This rapidly expanding legal market relies on third party Cannabis testing
laboratories to test products and determine whether or not they are safe for public consumption.
The standards for what is deemed ‘safe’ vary from state to state, as there is no federal input on
best testing practices for Cannabis, which is still classified as a Schedule I narcotic. Thus, a
patchwork of scientific testing protocols have emerged across legal Cannabis markets, where
analyses, analytes, limits of detection, and even product types can differ drastically between
regions.2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
In California, where legalization went into effect January 1st, 2018, every legal Cannabis
product must pass stringent testing requirements before it can be sold to consumers. In addition
to contaminant testing (e.g. pesticides, heavy metals, microbials, etc.), all Cannabis products
must be tested for the presence of six cannabinoids (∆ -THC, THCA, CBD, CBDA, CBN, CBG).3
Cannabinoids are a class of biologically active compounds that may induce a psychoactive and/or
medicinal effect on users.4,5 Commercially relevant cannabinoids, such as ∆9-
9
9
tetrahydrocannabinol (∆ -THC), cannabidiol (CBD), and cannabinol (CBN), are typified by two
core structural fragments: an oxygenated resorcinyl ring containing a pendant aliphatic chain, and
a p-menthyl ring with varied positions of unsaturation4,5 (vide infra). In the state of California, a
label claim containing accurate potency is required on all commercial products; if the cannabinoid
content of the product does not match the printed label claim within +/– 10%, the product must be
3
relabeled, or in some cases destroyed. To ensure accurate dosing information, third party testing
laboratories must maintain highly precise, accurate, and rugged analytical methods for a large
and ever-expanding number of Cannabis-infused matrices, with no standardized methods to
reference.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 4 of 20
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
With no prior literature on Cannabis testing, fractured scientific requirements from state to
state, and a constant stream of new product types, it falls on the third party Cannabis testing
laboratories to develop new methods for the analyses of these disparate matrices and establish
scientific standards for Cannabis product testing. Our investigation of potency testing on complex
Cannabis-infused matrices begins with Cannabis-infused chocolates, a ubiquitous product type
that accounted for 15% of retail sales in 2018 for the combined legal markets of California,
6
Washington, Oregon, and Arizona. Chocolate is a notoriously difficult food matrix for analyte
extraction and detection, as a high fat content and presence of polyphenolic compounds can
frustrate precise analytical testing.7–11 Relevant work from Khuda et al. showed that
immunodetection of allergens was complicated by matrix interference from components of dark
chocolate.7,8 In addition to fats and sugars added during chocolate processing, cocoa solids are
known to contain over seventy different organic flavoring compounds,12 which could also have
interactions with the little-studied cannabinoid analytes. The chemical complexity of the chocolate
matrix, combined with its omnipresence in legal Cannabis markets, made it an ideal product type
to begin our investigations.
Materials & Methods:
Analytical Materials: Milk and dark chocolate was supplied by Chill Chocolate (Oakland,
California, USA; 42% fat by weight). Cocoa powder was purchased from the Ghirardelli Chocolate
Company (San Leandro, California, USA; 100% unsweetened non-alkalized baking cocoa, 25%
9
9
fat by weight). ∆ -Tetrahydrocannabinol (∆ -THC) and cannabinol (CBN) were originally submitted
to Cw Analytical in oil form for quality assurance testing and used as submitted. Cannabidiol
(CBD) and cannabigerol (CBG) isolates were supplied by Vertosa Inc. (Oakland, California, USA).
All cannabinoids used in this study were tested by Cw Analytical and confirmed to be free of
9
pesticides, heavy metals, and residual solvents. CRM grade analytical standards of ∆ -THC, CBD,
CBN, and CBG were purchased from Cayman Chemical (Ann Arbor, Michigan, USA). An
ACS Paragon Plus Environment

Page 5 of 20
Journal of Agricultural and Food Chemistry
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
analytical standard of cannabidiol dimethyl ether (CBDD) was made by dissolving 100 mg of an
analytically pure fraction in 100 mL of HPLC-grade acetonitrile. Acetonitrile was purchased from
®
Emerald Scientific (San Luis Obispo, California, USA; ChemPure HPLC-grade). Methanol was
®
purchased from Emerald Scientific (TEDIA LC-MS grade). Dionized HPLC-grade water was
®
generated in-house utilizing a Millipore Milli-Q Gradient A10 water purification system. Formic
®
acid was purchased from Emerald Scientific (Sigma-Aldrich Suprapur 98–100%). Testing vials
were purchased from Nelson-Jameson (Marshfield, Wisconsin, USA; Capital Plastic Products
polypropylene 40 mL vials).
Synthetic Materials: All glassware was flame-dried prior to use. Dimethylformamide (DMF) was
degassed with argon and then passed through two 4 x 36 inch columns containing anhydrous
neutral A-2 alumina (8 x 14 mesh; LaRoche Chemicals; activated under a flow of argon at 350 ºC
for 12 hours) to remove residual H O. All other chemicals were purchased commercially and used
2
1
as received. NMR were recorded on a Bruker DRX-400 (400 MHz H) and CRYO-500 (125.7 MHz
13C). Proton chemical shifts are reported in ppm (δ) relative to internal tetramethylsilane (TMS, δ
0.00). Data are reported as follows: chemical shift (multiplicity [singlet (s), doublet (d), triplet (t),
triplet of doublets (td), multiplet (m), broad multiplet (br m)], integration, coupling constants [Hz],
part of the molecule). Carbon chemical shifts are reported in ppm (δ) relative to TMS with the
solvent resonance as the internal standard (CDCl , δ 77.16 ppm). NMR data were collected at 25
3
ºC. Analytical thin-layer chromatography (TLC) was performed with Silca Gel 60 Å F254
precoated plates (0.25 mm thickness). Flash chromatography was performed utilizing a Teledyne
Isco Combiflash® Rf+ automated flash chromatography system. High resolution mass
spectrometry was performed by the University of California, Irvine Mass Spectrometry Center.
Cannabinoid Stock Solutions: All cannabinoid stock solutions used in this study were made in
house at Cw Analytical. A concentration of 100 µg/mL in methanol was targeted for each stock
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 6 of 20
7
7
8
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
solution; this value is indicative of the concentration measured when testing Cannabis-infused
edible products. Each stock solution was made in batches of 4 L at a time. All experiments for a
given cannabinoid were performed using a single batch of stock solution. Each stock solution was
tested [n = 10] for potency and the values averaged to determine the actual concentration. When
not in use, stock solutions were stored at 4 ºC and warmed to room temperature before use.
Chocolate Testing Procedure: Chocolate was blended in a food processor until it became a fine
powder, roughly 1 minute. Homogenized chocolate was then weighed into test vials in either 1 g,
2 g, or 3 g amounts, with [n = 10] for each and with no sample exceeding +/– 1% of the given
testing value. An aliquot of 20 mL of cannabinoid stock solution in methanol was then added to
®
each vial from a Dispensette S brand bottle top dispenser, to serve as both cannabinoid delivery
phase and extraction solvent. Vials were vortexed for 30 seconds, sonicated at 2200 W for 20
minutes, centrifuged at 1500 rpm for 5 minutes, and winterized at –20 ºC for 30 minutes. The
solution was filtered (0.45 µm PTFE) into ALS vials and analyzed for cannabinoid content via
HPLC-DAD. Recovery percentage was determined by dividing the average [n = 10] concentration
of analyte (µg/mL) for a given amount of chocolate product by the average [n = 10] concentration
of analyte (µg/mL) in the blank stock solution.
HPLC Method: HPLC separation was performed on an Agilent 1260 Infinity II using a Restek
Raptor ARC-18 column (150 x 4.6 mm inner diameter, 2.7 µm particle, and 90 Å pore size) at 40
ºC. Mobile phases were 0.1% formic acid in HPLC-grade water [mobile phase A (MPA)] and 0.1%
formic acid in HPLC-grade acetonitrile [mobile phase B (MPB)] at a flow rate of 1.5 mL/min on a
gradient of 75% MPB for minutes 0–3, then 75%–100% MPB over minutes 3–7. The injection
volume was 2.0 µL. The detector was an Agilent 1260 DAD HS measuring a signal wavelength
of 218 nm. All results were quantitated against a five-point external calibration curve. All
1
1
1
00
01
02
ACS Paragon Plus Environment

Page 7 of 20
Journal of Agricultural and Food Chemistry
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
calibration curves were linear-fit, set to include the origin, with a weighting factor of 1/x and
2
correlation coefficients (R ) of >0.999.
Synthesis of cannabidiol dimethyl ether (CBDD): CBDD was synthesized according to a modified
procedure from Mechoulam.13 CBD (5.00 g, 15.9 mmol) was added to a 250 mL round bottom
flask equipped with stir bar, and dissolved in anhydrous DMF (90 mL). Next, K CO (11.8 g, 85.9
2
3
mmol) and iodomethane (3.6 mL, 59 mmol) were added and allowed to stir for 24 h at room
temperature under a flow of N . During the course of the reaction, the solution changed color from
2
a deep purple hue to golden yellow. After 24 h, deionized water was added to the flask (150 mL)
and the solution extracted with diethyl ether (3 x 150 mL). The organic layers were combined and
washed with a saturated solution of NaCl (150 mL), dried over MgSO , and filtered. The organic
4
solvent was removed via rotary evaporation, and the crude product was purified via flash column
chromatography (5% diethyl ether/hexanes) to afford the title compound as a clear, pale yellow
1
oil (4.25 g, 12.4 mmol, 78%). TLC R = 0.8 (5% diethyl ether/hexanes); H NMR (400 MHz, CDCl )
f
3
δ 6.34 (s, 2, C H ), 5.23 (s, 1, CHCH=C(CH )(CH )), 4.46–4.42 (m, 2, CH =C(CH )(CH)), 3.99 (m,
6
2
3
2
2
3
1, (CH) CHAr), 3.74 (s, 6, OCH ), 2.90 (td, 1, J = 10.7, 4.1, (CH ) HC=CH ), 2.54 (t, 2, J = 7.8,
2
3
2 2
2
ArCH CH ), 2.25–2.14 (br m, 1, CH CH C=C), 1.98 (m, 1, CH CH C=C), 1.67–1.60 (br m, 10,
2
2
2
2
2
2
CH CH CH CH ), 1.34 (m, 4, CH CH CH ), 0.91 (t, 3, J = 7.0, CH CH ); 13C NMR (125.7 MHz,
2
2
2
3
2
2
2
2
3
CDCl ) δ 158.9 (2C), 149.60, 141.97, 131.23, 126.11, 119.10, 109.72, 105.10 (2C), 56.03 (2C),
3
45.34, 36.55, 36.25, 31.84, 31.14, 30.89, 29.85, 23.56, 22.70, 19.18, 14.19; HRMS (TOF MS CI+)
+
m/z calcd for C H O H [M + H] 343.2637; found 343.2632.
23
34
2
Results & Discussion:
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 8 of 20
125
9
1
1
1
1
1
1
26
27
28
29
30
31
Figure 1: Investigation of matrix effects between various chocolate types and A) ∆ -
9
tetrahydrocannabinol (∆ -THC), B) cannabidiol (CBD), with chemical structures shown below.
Blue squares represent milk chocolate, orange triangles represent dark chocolate, and grey
circles represent cocoa powder.
Our investigation of potential matrix effects from chocolate centered on three different
common chocolate sources: milk chocolate, dark chocolate, and unsweetened non-Dutch
process cocoa powder. Milk and dark chocolates have very similar chemical compositions, but
differ in the amount of cocoa solids and added milk fats. Conversely, cocoa powder has a minimal
132
133
134
135
136
137
138
139
140
141
14
amount of fat present, and is mostly cocoa solids by weight. In hopes of testing a purer source
of cocoa, the cocoa used in this experiment has not undergone the ‘Dutch process’, a post-
production step where alkaline compounds are added to preserve color and flavor. These three
9
9
chocolate types were first subjected to a stock solution of ∆ -tetrahydrocannabinol (∆ -THC), a
ubiquitous cannabinoid in legal Cannabis edible products. For each chocolate type, three different
sample loadings were tested (one gram, two grams, three grams chocolate, [n = 10] replicates)
and plotted against the cannabinoid concentration, as measured by HPLC analysis. It is
ACS Paragon Plus Environment

Page 9 of 20
Journal of Agricultural and Food Chemistry
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
hypothesized that if the recovery rate of a cannabinoid is lower at higher sample loadings, it might
be indicative of matrix interference from the chocolate.
9
After subjecting the three chocolate types to a stock solution of ∆ -THC, a clear trend of
signal suppression emerges (Figure 1A). Both milk and dark chocolate show high recovery rates
with 1 g loadings, but then decrease linearly as more chocolate is added. Cocoa powder displays
9
a similar correlation between higher sample loadings and lower recovery rates of ∆ -THC, but to
a lesser degree than milk and dark chocolates. This initial experimentation suggests that the
presence of chocolate may inhibit the complete recovery of cannabinoids in solution, presenting
implications for finished product testing. To determine the scope of this potential matrix effect, we
turned our focus to cannabidiol (CBD), another common cannabinoid in commercial Cannabis
products. After performing the same recovery experiment with a stock solution of CBD, the
negative correlation between sample loading and recovery rates of CBD was not as pronounced
9
as with ∆ -THC (Figure 1B). Neither milk chocolate nor dark chocolate had substantially
diminished recoveries of CBD, and all permutations yielded recovery rates >98%, compared to
9
~93% ∆ -THC recovery for milk and dark chocolates. While this difference in signal suppression
9
is relatively subtle, it is surprising due to the structural similarity between ∆ -THC and CBD
9
(Scheme 1). CBD contains two phenolic -OH groups on the resorcinyl moiety, whereas ∆ -THC
has undergone electrophilic cyclization between the pendant allyl group and one of the phenolic
-OH groups. We hypothesized that this difference in functional groups may account for the
differing magnitudes of signal suppression seen in Figure 1.
162
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 10 of 20
1
1
1
1
63
64
65
66
Scheme 1: The structures of four common biogenic cannabinoids with p-menthyl and resorcinyl
fragments annotated; phenolic -OH groups are outlined in red.
To test the hypothesis that structural features of cannabinoids can affect signal
suppression rates, the recovery experiment was again repeated with two additional cannabinoids:
cannabinol (CBN) and cannabigerol (CBG). While not commonly found in large quantities in edible
Cannabis products, their varied structures can provide further insight into the relationship between
167
168
169
170
171
172
173
9
cannabinoid structure and chocolate matrix effect. CBN is a derivative of ∆ -THC that has
9
undergone oxidative aromatization of the p-menthyl moiety, and like ∆ -THC only contains one
5
9
phenolic -OH group. CBG is a biogenic precursor to both ∆ -THC and CBD, and features a long,
linear isoprenyl residue and two free phenolic -OH groups, similar to CBD (see Scheme 1).
174
175
176
177
178
179
Figure 2: Investigation of matrix effects between various chocolate types and A) cannabinol
(CBN), B) cannabigerol (CBG), with chemical structures shown below. Blue squares represent
milk chocolate, orange triangles represent dark chocolate, and grey circles represent cocoa
powder.
1
80
81
The results of the recovery experiments with CBN and CBG are seen in Figure 2. The
recovery rates of CBN in the presence of all chocolate types decrease as the amount of chocolate
1
ACS Paragon Plus Environment

Page 11 of 20
Journal of Agricultural and Food Chemistry
9
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
increases (Figure 2A), which roughly aligns with the trends seen for ∆ -THC, which also contains
only one phenolic -OH group. It is noted that the presence of cocoa powder substantially
decreased the recovery of CBN, which suggests that the aromatized p-methyl moiety may cause
additional interactions between the cannabinoid and cocoa solids (vide infra). Conversely, the
recovery rates of CBG remained very high for all chocolate types at all sample loadings (Figure
2B). Similar to CBD, all recoveries for CBG were above 98%, which suggests that CBG and CBD
have similar chemical interactions with the chocolate matrix in solution.
Viewed as a whole, the data in Figures 1 and 2 support the hypothesis that the intensity
of chocolate matrix interference is related to structural features of the cannabinoids, specifically
the number of phenolic -OH groups. Cannabinoids containing two phenolic -OH groups (e.g. CBD,
CBG) result in a subtle matrix effect, and high recovery rates upon analysis. A cannabinoid with
9
only a single phenolic -OH group (e.g. ∆ -THC, CBN) will have more substantial matrix
interference, with a concomitant drop in recovery rate. The pronounced effect of phenolic -OH
moieties can likely be attributed to a change in equilibrium between the polar, protic methanol
solvent and the less polar fat-rich chocolate matrix. Cannabinoids with two polar phenolic -OH
groups will have an equilibrium favoring the solvent, and thus have higher recovery rates. A
cannabinoid containing a single phenolic -OH functional group is less preferentially solvated by
methanol, and thus at equilibrium will be distributed within both the chocolate and solvent phases,
resulting in enhanced matrix interference and lower recovery rates. Extrapolating from these
findings, a cannabinoid with no phenolic -OH groups could be expected to exhibit substantially
lower recovery rates than even the singly phenolic -OH cannabinoids, as the lack of a polar
functional group would drastically shift the equilibrium partitioning in favor of the chocolate matrix.
Testing a cannabinoid without phenolic -OH groups could provide strong evidence in favor of this
hypothesis.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 12 of 20
2
06
07
2
Scheme 2: Synthesis of cannabidiol dimethyl ether (CBDD) from cannabidiol (CBD).
208
209
210
211
212
213
214
To determine if a cannabinoid without phenolic -OH groups would exhibit a strong matrix
13
effect, a synthetic modification was made to CBD (Scheme 2). By methylating the two phenolic
-OH moieties of CBD, cannabidiol is converted to cannabidiol dimethyl ether (CBDD), a non-
biogenic cannabinoid that has been shown to stimulate weight gain in (ApoE)-deficient BALB/c.
KOR/Stm Slc-Apoeshl mice. CBDD contains no phenolic -OH groups, and thus direct comparison
to CBD can elucidate the importance of free phenolic -OH functional groups on cannabinoid
recovery rates.
15
215
216
217
218
219
220
Figure 3: Investigation of matrix effects between various chocolate types and cannabidiol
dimethyl ether (CBDD). Blue squares represent milk chocolate, orange triangles represent dark
chocolate, and grey circles represent cocoa powder.
Consistent with our hypothesis, CBDD displayed a high degree of signal suppression
9
221
(Figure 3). Similar to ∆ -THC and CBN, the two cannabinoids containing one phenolic -OH group,
ACS Paragon Plus Environment

Page 13 of 20
Journal of Agricultural and Food Chemistry
222
223
224
225
226
227
228
the decrease in recovery rates shows strong linearity, and is directly correlated to the amount of
chocolate present. However, the intensity of CBDD signal suppression is substantially higher than
9
that of either ∆ -THC or CBN, consistent with the hypothesis that structural features of
cannabinoids play a significant role in determining if recovery rates will be altered by matrix
effects.
A)
2
29
30
2
B)
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 14 of 20
2
31
32
2
Figure 4: Percent recovery for five cannabinoids from A) milk chocolate and B) dark chocolate.
Blue circles represent ∆ -THC, orange squares represent CBD, grey triangles represent CBN,
yellow diamonds represent CBG, and blue crosses represent CBDD.
9
233
234
235
236
When the data from Figures 1–3 are plotted as percent recovery, the importance of
cannabinoid structure on recovery rate is made clear (Figure 4). For both milk and dark chocolate
matrices, analytes with two phenolic -OH groups (CBD, CBG) have percent recoveries >98% for
237
238
239
240
241
242
243
244
245
246
247
248
249
9
all quantities of chocolate tested. ∆ -THC and CBN, which both contain one phenolic -OH group,
behave similarly to one another, with recovery rates of 95% at 2 g chocolate added, and ~93% at
3 g chocolate. Synthetic CBDD, which bears no phenolic -OH group, exhibits dramatically lower
recovery rates than CBD. The 88% recovery seen with 2 g milk chocolate and 85% recovery with
2 g dark chocolate are lower than the lowest recoveries seen on either monophenolic
cannabinoid, and the percent recovery dropped <80% with 3 g dark chocolate loading.
The data presented in this study establishes for the first time an interaction between the
chocolate matrix and cannabinoid analytes that can impact finished product potency testing. The
degree to which the chocolate matrix interferes with cannabinoid analysis depends on two factors:
the amount of chocolate present and the chemical structure of the cannabinoid analyte. The
amount of chocolate present during testing is directly proportional to the degree of interference,
ACS Paragon Plus Environment

Page 15 of 20
Journal of Agricultural and Food Chemistry
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
as the more chocolate present in the sample the lower the percent recovery of analyte. For each
combination of chocolate matrix and analyte, the 3 g sample loading afforded the lowest percent
recovery, followed by the 2 g loading, with 1 g of chocolate providing the highest percent recovery.
This near linearity of this trend suggests that when present in large quantities, some component
of the chocolate matrix alters the equilibrium state of the cannabinoid analyte, increasing
partitioning within the chocolate matrix. This shift in equilibrium away from the liquid phase results
in less cannabinoids in solution, and thus lower recovery rates.
The second major factor that dictates the magnitude of matrix interference is the
cannabinoid’s chemical structure. Despite the general structural similarities common to all
cannabinoids, the effect of the phenolic -OH groups on analyte recovery is striking. If there are
two phenolic -OH groups present on the resorcinyl fragment of a cannabinoid, there will be
negligible interference from the chocolate matrix (recovery rates >98%). Monophenolic
cannabinoids suffer from minor matrix interference (recovery rates >90%), and CBDD, which
contains no phenolic -OH groups, has substantial interference from the chocolate matrix (recovery
rates >75%). The pronounced effect of phenolic -OH moieties on analyte recovery rates likely
stems from the analyte’s equilibrium partitioning between liquid and solid phases. Compounds
with more phenolic -OH groups are more polar, and likely have increased solubility in methanol,
a polar protic solvent. An equilibrium favoring the liquid phase would result in higher recovery
rates, such as those observed for both CBD and CBG (Figure 4). By decreasing the amount of
phenolic -OH groups, there are fewer favorable interactions with the solvent, and thus equilibrium
is expected to shift away from the liquid phase and towards the chocolate matrix. A cannabinoid
without phenolic -OH moieties has substantially higher non-polar character, and is likely to have
increased solubility in non-polar matrices. This decreased solubility would correlate with low
recovery rates, as seen with CBDD in Figures 3 & 4.
Increasing non-polar character in cannabinoids would not only impact the solubility of the
analyte in solution, but is also likely to increase interactions with the chocolate matrix. In addition
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 16 of 20
276
277
278
279
280
281
282
283
284
285
286
to added sugars and naturally occurring cocoa solids, chocolate is high in fat, with both milk and
dark chocolates tested containing approximately 42% total fat by weight. Cannabinoids are known
to be lipophilic and have a high degree of solubility in fats, a characteristic corroborated by the
centuries old practice of extracting Cannabis with butter or ghee,16 and the long-term
17
accumulation of cannabinoids seen in human fatty tissue. Thus, it is expected that cannabinoids
dissolved in solution could be attracted to the chocolate matrix, due at least in part to its high fat
content. This lipophilic interaction would be strongest for non-polar cannabinoids (i.e. containing
no phenolic -OH groups) and weakest with polar cannabinoids (i.e. containing one or multiple
phenolic -OH groups). This proposed interaction would result in lower recovery rates for non-polar
cannabinoids compared to polar cannabinoids, a trend that is supported by the data in Figure 4,
where CBDD has a substantially lower recovery rate than CBD when in the presence of chocolate.
287
288
289
290
291
Figure 5: Percent recovery for five cannabinoids from cocoa powder. Blue circles represent ∆9-
THC, orange squares represent CBD, grey triangles represent CBN, yellow diamonds represent
CBG, and blue crosses represent CBDD.
2
92
93
Our hypothesis that lipophilic interactions are at the heart of chocolate matrix interference
is further bolstered by the data collected with the cocoa powder matrix (Figure 5). In almost all
2
ACS Paragon Plus Environment

Page 17 of 20
Journal of Agricultural and Food Chemistry
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
instances, cannabinoid recovery rates from the cocoa powder matrix were higher than the
analogous recovery rates from milk and dark chocolates. Unsweetened non-alkalized cocoa
powder is pressed from raw cocoa liquor and thus is mostly cocoa solids, with a lower fat content
than finished chocolate products.12,14 The cocoa powder tested in this study was 25% fat by
weight, compared to 42% fat by weight for the milk and dark chocolates. This comparatively lower
fat content means there are less lipophilic interactions between the cannabinoid analytes and the
cocoa matrix, which in turn would explain the higher recovery rates observed for cocoa powder
(see Figures 1–3).
It is noted that this trend is valid for all cannabinoids except for CBN, which exhibits an
unusually low recovery rate when extracting from the cocoa powder matrix. Even at 1 g of cocoa
9
powder, recovery rates of CBN are lower than that of ∆ -THC at 3 g cocoa powder. This sets it
apart from the other four tested analytes, where the data matches the overall trends seen in Figure
9
4 (i.e. CBD/CBG recovery > ∆ -THC recovery >> CBDD recovery). This suggests that there may
be another chemical interaction between cannabinoids and the chocolate matrix at play, one that
may be related to the overall amount of cocoa solids present and/or unique structural features of
CBN. Specifically, CBN is the only tested cannabinoid with a fully aromatized p-menthyl fragment,
which in turn makes the molecular skeleton of CBN almost entirely planar. Cocoa solids extracted
from cocoa beans are rich with flavonoids, such as (+)-catechin, (–)-epicatechin, and (+)-
gallocatechin, and on average can contain up to 8% flavonoids by weight.12,18 These naturally
occurring compounds are classified as flavan-3-ols, a class of molecules that have been shown
to inhibit proteins via non-covalent London interactions between non-polar polarizable aromatic
rings.18 It is hypothesized that an analogous non-covalent London interaction between naturally
occurring flavan-3-ols in cocoa and the highly π-conjugated CBN may account for the high degree
of matrix interference observed between CBN and the cocoa matrix. Experiments testing this
hypothesis are ongoing.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 18 of 20
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
The results disclosed in this manuscript represent the early stages of modern Cannabis
research, and underscore the need for further scientific investigation in the field of Cannabis
analysis. In analogous fields such as pharmaceuticals, food, and agriculture testing, matrix effects
and analyte suppression are well studied, leading to highly precise and accurate analyses. For
Cannabis analytical testing, such research is scant, and specific molecular interactions that may
affect precise testing are either understudied or presently unknown. Matrix interference from
chocolate products on cannabinoid analytes appears to be quite nuanced, and has been shown
to be dependent on quantity of chocolate analyzed, composition with respect to fat and cocoa
solids, and multiple structural features of the cannabinoid analytes. Many of these factors are not
limited to just chocolate matrices, as baked goods and topical products are high in fat and flavoring
additives, and thus might interfere with analysis of cannabinoid content. Any scientific efforts to
standardize Cannabis analytical methods must be based on detailed studies of the molecules and
matrices involved. Further scientific advances in Cannabis testing will be required in order for the
Cannabis industry to continue to make strides away from the black market and for the long-term
stability of the Cannabis industry as a whole.
Abbreviations:
9
9
9
∆ -THC, ∆ -tetrahydrocannabinol; THCA, ∆ -tetrahydrocannabinolic acid; CBD, cannabidiol;
CBDA, cannabidiolic acid; CBN, cannabinol; CBG, cannabigerol; HPLC, high pressure liquid
chromatography; DMF, dimethylformamide; TMS, trimethylsilane; TLC, thin-layer
chromatography; CBDD, cannabidiol dimethyl ether.
Acknowledgements:
We gratefully acknowledge Dr. Elizabeth Jarvo of the University of California Irvine for
collaboration on the synthesis of CBDD, and for her ongoing mentorship. Also, we thank
ACS Paragon Plus Environment

Page 19 of 20
Journal of Agricultural and Food Chemistry
344
345
346
Vertosa of Oakland, CA for their generous donation of CBD and CBG. Finally, we wish to thank
Chill Chocolate of Oakland, CA for supplying chocolate throughout the course of this study.
347
348
349
350
351
352
Funding Sources: The authors declare no competing financial interest.
9
Supporting Information: HPLC chromatograms of ∆ -THC, CBD, CBN, CBG, and CBDD
prepped with 3 g of dark chocolate (Figures SI1–SI5) (PDF). The supporting information is
available free of charge on the ACS Publications website at DOI: XXXXXXXXXXX.
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
References:
1. State Medical Marijuana Laws. https://www.ncsl.org/research/health/state-medical-
marijuana-laws.aspx (last access Feburary 17th, 2020).
2. Labs, W. Cannabis testing is an exact science - regulations are not.
https://www.foodengineeringmag.com/articles/98370-cannabis-testing-is-an-exact-
science---regulations-are-not (last access February 17th, 2020).
3. Bureau of Cannabis Control Text of Regulations.
https://bcc.ca.gov/law_regs/cannabis_order_of_adoption.pdf (last access February 17th
2020).
4. ElSohly, M. A.; Slade, D. Chemical constituents of marijuana: The complex mixture of
natural cannabinoids. Life Sci. 2005, 78, 539–548.
5. Hanuš, L. O.; Meyer, S. M.; Muñoz, E.; Taglialatela-Scafati, O.; Appendino, G.
Phytocannabinoids: a unified critical inventory. Nat. Prod. Rep. 2016, 33, 1357–1392.
6. Thompson, A. Which types of cannabis confectionary edibles are most popular? PMCA
event offers insight. https://www.candyindustry.com/articles/88908-which-types-of-
cannabis-confectionery-edibles-are-most-popular-pmca-event-offers-insight (last access
April 5th, 2020).
7. Khuda, S. E.; Jackson, L. S.; Fu, T.-J.; Williams, K. M. Effects of processing on the
recovery of food allergens from a model dark chocolate matrix. Food Chem. 2015, 168,
580–587.
8. Khuda, S.; Slate, A.; Pereira, M.; Al-Taher, F.; Jackson, L.; Diaz-Amigo, C.; Bigley, III, E.
C.; Whitaker, T.; Williams, K. Effect of Processing on Recovery and Variability
Associated with Immunochemical Analytical Methods for Multiple Allergens in a Single
Matrix: Dark Chocolate. J. Agric. Food Chem. 2012, 60, 4204–4211.
9. Khuda, S. E.; Williams, K. M.; Effect of Processing on Dark Chocolate Composition: A
Focus on Allergens. In Processing and Impact on Active Components in Food, 1st
edition; Preedy, V., Ed.; Academic Press, Cambridge, MA, 2015; 667–674.
10. Shefcheck, K. J.; Callahan, J. H.; Musser, S. M. Confirmation of Peanut Protein Using
Peptide Markers in Dark Chocolate Using Liquid Chromatography-Tandem Mass
Spectrometry (LC-MS/MS). J. Agric. Food. Chem. 2006, 54, 7953–7959.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 20 of 20
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
11. Taylor, S. L.; Nordlee, J. A.; Niemann, L. M.; Lambrecht, D. M. Allergen immunoassays -
considerations for use of naturally incurred standards. Anal. Bioanal. Chem. 2009, 395,
83–92.
12. Afoakwa, E. O. The chemistry of flavour development during cocoa processing and
chocolate manufacture. In Chocolate Science and Technology, 2nd edition; John Wiley
& Sons, Ltd., West Sussex, United Kingdom, 2016; 159–170.
13. Tchilibon, S.; Mechoulam, R. Synthesis of a Primary Metabolite of Cannabidiol. Org.
Lett. 2000, 2, 3301–3303.
14. McGee, H. Sugars, Chocolate, and Confectionery. In On Food And Cooking: The
Science And Lore Of The Kitchen, 1st edition; Scribner, New York, NY, 2004; 645–712.
15. Takeda, S.; Hirota, R.; Teradaira, Takeda-Imoto, M.; Watanabe, K.; Toda, A.; Aramaki,
H. Cannabidiol-2’,6’-dimethyl ether stimulates body weight gain in apolipoprotein E-
shl
deficient BALB/c. KOR/Stm Slc-Apoe mice. J. Toxicol. Sci. 2015, 40, 739–743.
16. Drake, B. A Natural, Inexpensive High. In The Marijuana Food Handbook, 2nd edition;
Ronin Publishing, Oakland, CA, 2002; 17–23.
17. Ashton, C. H. Pharmacology and effects of cannabis: a brief review. Br. J. Psychiatry
2001, 178, 101–106.
18. Bordenave, N.; Hamaker, B. R.; Ferruzzi, M. G. Nature and consequences of non-
covalent interactions between flavonoids and macronutrients in foods. Food Funct. 2014,
5, 18–34.
404
405
406
Table of Contents Graphic:
407
ACS Paragon Plus Environment