Here's a couple of different choices inspired by LanX's Re: Help with parsing a file (hash keys are unique)

#!/usr/bin/perl

use strict; # https://perlmonks.org/?node_id=11144249
use warnings;

my $file = 'data.11144249';
open my $fh, '<', $file or die "$! opening $file";
local $_ = do{ local $/; <$fh> }; # slurp entire file

my %solvent_face;
# as requested
$solvent_face{ "$3_$1" } = $2 while
  /\b (F\S+) \h+ ([-\d.]+) (?=.*? (\S+) \h+ 0.00000000 )/gsx;
# by solvent, then arrays of each F***
push @{ $solvent_face{ $3 }{ $1 } }, $2 while
  /\b (F\S+) \h+ ([-\d.]+) (?=.*? (\S+) \h+ 0.00000000 )/gsx;
# by F***, then arrays of solvent
push @{ $solvent_face{ $1 }{ $3 } }, $2 while
  /\b (F\S+) \h+ ([-\d.]+) (?=.*? (\S+) \h+ 0.00000000 )/gsx;
use Data::Dump 'dd'; dd \%solvent_face;
[download]

Outputs:

{
  "acetonitrile" => {
    "F002"  => [1.69945785],
    "F011"  => [0.74578421],
    "F101"  => [2.67268035],
    "F11-1" => [1.64808218],
    "F111"  => [1.95840198],
  },
  "acetonitrile_F002" => 1.69945785,
  "acetonitrile_F011" => 0.74578421,
  "acetonitrile_F101" => 2.67268035,
  "acetonitrile_F11-1" => 1.64808218,
  "acetonitrile_F111" => 1.95840198,
  "F002" => { acetonitrile => [1.69945785], h2o => [4.66656083] },
  "F011" => { acetonitrile => [0.74578421], h2o => [26.13597035] },
  "F101" => { acetonitrile => [2.67268035], h2o => [32.47411476] },
  "F11-1" => { acetonitrile => [1.64808218], h2o => [29.58963453] },
  "F111" => { acetonitrile => [1.95840198], h2o => [30.24092207] },
  "h2o" => {
    "F002"  => [4.66656083],
    "F011"  => [26.13597035],
    "F101"  => [32.47411476],
    "F11-1" => [29.58963453],
    "F111"  => [30.24092207],
  },
  "h2o_F002" => 4.66656083,
  "h2o_F011" => 26.13597035,
  "h2o_F101" => 32.47411476,
  "h2o_F11-1" => 29.58963453,
  "h2o_F111" => 30.24092207,
}
[download]